JP3499672B2 - Automatic performance device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic musical instrument, and more particularly to a function for ending a musical performance.

[0002]

2. Description of the Related Art Conventionally, when ending a performance in an automatic accompaniment device, in addition to a normal performance pattern, a performance pattern (ending pattern) that reminds the end of the performance is stored in advance and the ending pattern is played. The performance is ended by Koyo. This ending pattern is stored in the same form as a normal performance pattern, and only the musical content of the performance pattern is different.

[0003]

In the conventional apparatus, it is necessary to prepare not only a normal performance pattern but also an ending pattern separately at the end of the performance, and an extra storage area in the memory is required for this ending pattern. Will be needed. Therefore, an object of the present invention is to effectively end the performance of the automatic accompaniment with a simple structure without consuming memory for storing the ending pattern.

[0004]

In order to solve the above-mentioned problems, the automatic performance device according to claim 1 of the present invention uses a performance pattern that is different from the ending pattern and that includes pitch information. A storage means for storing a plurality of pieces of music corresponding to the above, a performance means for automatically playing a performance pattern corresponding to a style from the storage means, a performance end instruction means for instructing the end of performance, and a performance by the performance end instruction means. After the end instruction, the head portion of a predetermined performance pattern corresponding to the style is read from the storage means, and the musical tone stored at the head portion is reproduced according to the pitch information included in the predetermined performance pattern. And a performance control means for controlling the performance means so as to end the performance after continuing the sounding for a predetermined time not depending on the predetermined performance pattern. Also, the automatic performance device according to claim 2 of the present invention is a performance pattern different from the ending pattern, a storage means for storing a plurality of performance patterns including pitch information corresponding to styles, and a code designating a code. Designating means, performance means for reading a performance pattern corresponding to a style from the storage means, converting pitch information in accordance with the designated chord to automatically perform performance, and performance end instruction means for instructing performance end Then, after the performance end instruction means gives an instruction to end the performance, the head portion of a predetermined performance pattern corresponding to the style is read from the storage means,
The tone pitch stored in the head portion is the pitch obtained by converting the pitch information contained in the predetermined performance pattern according to the chord designated by the chord designating means, and does not depend on the predetermined performance pattern. And a performance control means for controlling the performance means so as to end the performance after continuing the sound generation for a predetermined time.

With such a configuration, when a performance is finished, it is based on a normal performance pattern (such as a normal pattern) even if the memory does not have a special performance pattern (ending pattern) for giving a feeling of end. , It is possible to generate a musical sound that evokes the end of the performance, and thus it is possible to realize the end of the performance with a feeling of end with a simple configuration.

As a musical tone suitable for recalling the end of the performance, the musical tone indicated by (*) in FIG. 1 (A) is shown in FIG.
Generally, a musical tone that is generated at the timing of the beginning of the performance pattern, such as the pronunciation as shown in (B), is effective, and if the ending process is performed with this musical tone, a more preferable ending feeling may be obtained. Many.

In addition to the first performance pattern for a normal performance, the second performance pattern is stored in the storage means.
The performance control means prepares the performance pattern, and when the performance end instruction means gives the performance end instruction, the performance control means automatically plays the second performance pattern, and then ends the performance from the first performance pattern. The playing means may be configured so that a musical tone suitable for recalling is read out, the musical tone is sounded for a predetermined length that does not depend on the length of the musical tone, and then the playing is ended.

As described above, a more complicated ending can be realized by inserting another second performance pattern before ending the performance based on the musical tone of the first performance pattern.

The first performance pattern can be a normal pattern, and the second performance pattern can be a fill-in pattern. By using the fill-in pattern as the second performance pattern in this way, it is possible to obtain a more effective end feeling in the performance.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 shows an automatic musical instrument as an embodiment of the present invention. In this embodiment, the present invention is applied to an automatic accompaniment function in an automatic performance device.

A CPU (Central Processing Unit) 1 controls the entire apparatus and operates based on a program stored in a program memory 10. The tempo clock generator 2 generates a clock for advancing the automatic accompaniment, and changes the clock cycle according to the tempo instructed by the CPU 1. The sound source 3 emits / silences a musical tone signal in response to an instruction from the CPU 1 and changes a tone color to be generated, and the generated musical tone signal is emitted through the amplifier 11 and the speaker 12.

The display unit 4 is a device for displaying the tempo of this device, a code detected by key depression, and the like. The operator group 5 includes various operators operated by the user, and as shown in FIG. 4, an automatic performance start / stop switch 51, a fill-in switch 52, an ending switch 53, a tempo dial 54, and a style selection switch 55.
And so on. The keyboard 6 is for operating the performance,
In this embodiment, the area near the center of the keyboard is divided into left and right areas, which are used as a melody playing area and a chord designating area, respectively.

The accompaniment data memory 7 is a ROM that stores performance data for accompaniment. The alternation table 8 is a ROM that stores a table used when the accompaniment data is reread according to the chord specified on the keyboard 6.
Is. The work area 9 is composed of a RAM and includes various registers.

FIG. 3 shows a configuration example of accompaniment data stored in the accompaniment data memory 7. The accompaniment data stores a plurality of performance forms for each accompaniment style. Here, there are rock, pop, jazz, etc. as the accompaniment style,
There are patterns such as intro, normal, and fill-in as a plurality of playing modes.

Here, the intro pattern is a pattern for recalling the start of the performance, the normal pattern is a basic pattern for performing a normal performance and is repeatedly played, and the fill-in pattern is It is an improvisational pattern that is decorated by filling the blank part of the melody pattern and is used temporarily.

Each accompaniment data has a pattern length at the beginning and is followed by a sequence of event data. The pattern length is the number of clocks that the pattern is played. The event data is instruction data for the sound source 3, such as sound generation and tone color change, and is stored in any of the following formats.

1. Timing: Part: [Note on]: Note number 2. Timing: Part: [Note Off]: Note number 3. Timing: Part: [Tone change]: Tone number

Here, the timing indicates the timing of instructing note-on, note-off, etc., and is expressed by the number of clocks when the pattern head is 0. The part is a chord part, a bass part, or a drum part. The note number represents the pitch of the sound to be sounded / muted. Note that note-on and note-off mean sounding and mute, respectively.

FIG. 4 shows details of the operator group 5. The START / STOP switch 51 is a switch for instructing the start / stop of the automatic accompaniment. It is composed of a toggle switch. Press once to start, and press it again to stop. Switch. A fill-in switch 52 is a switch for inserting a fill-in pattern. The ending switch 53 is a switch for instructing to end the performance by playing a performance pattern that reminds the end of the performance. The style (STYLE) selection switch 55 is a switch group for designating an accompaniment style, and at the time of automatic accompaniment, the data of the accompaniment pattern designated by these switches is played. The tempo (TEMPO) dial 54 is a dial for designating an accompaniment tempo, and the designated tempo is from the CPU 1 to the tempo clock generator 2.
The tempo clock generator 2 generates a clock according to the instructed tempo.

The operation of the embodiment apparatus will be described below with reference to the flow charts of FIGS. CPU 1 in FIG.
The flowchart of the main routine that is processed by is shown. When the power of the apparatus of the embodiment is turned on, initialization is performed (step A). In this initial setting, the work area is initialized and the clock interrupt from the tempo clock generator 2 is set to be disabled (that is, the automatic accompaniment stop state is set).
And so on.

Following the initial setting, manipulator processing is performed (step B). In this manipulator process, as will be described later in detail with reference to the flowchart of FIG. 6, the contents of the register on the work area 9 are appropriately changed according to the operation of each manipulator.

Following the manipulator processing, key depression processing is performed (step C). In this key-depression process, as will be described later in detail with reference to the flowchart of FIG. 7, the tone generator 3 is instructed to generate a musical tone in response to the operation of the keyboard, and the code designated by the key-depression is stored in the register.

After the manipulator processing, display processing is performed (step D). In this display processing, various information such as tempo value and designated code is displayed on the display section 4. When the display processing is completed, the operation processing (step B) is returned to, and thereafter, the operation processing, the key depression processing (step C), and the display processing (step D) are sequentially repeated.

In the apparatus of this embodiment, in addition to the processing by this main routine, every time the tempo clock generator 2 generates a tempo clock, the tempo clock interrupt processing routine shown in FIGS. To be executed.

Before describing the flowcharts of FIGS. 6 to 10, various registers used in these processes will be described first.

The tempo clock “clock” is a register for counting the tempo clock. The pattern length "length" is a register for setting the pattern length of accompaniment data and comparing it with the tempo clock.

The status “status” is a register indicating a performance form, and indicates a pattern to which the current accompaniment pattern is transferred after the performance is finished.
Takes a value of ~ 4. The meaning of each value is as follows. "0": Normal playing mode (indicating that the next shift is to the normal pattern) "1": This value is set when the fill-in switch 52 is pressed. As a result, the next shift is made to the fill-in pattern, and when the shift is made, "status" = "0" is set. "2": This value is set when the ending switch 53 is pressed. Next move to fill pattern
"status" is set to "3". "3": Next, the process shifts to the normal pattern, and the process of maintaining the sound of the beginning of the pattern for a predetermined time is performed. In this embodiment, the predetermined time for maintaining the pronunciation is half of one measure (two beats if one measure is 4 beats). “4”: Continuous sound (sound that keeps sounding) is being generated. There is no next pattern (that is, the performance ends).

The mask flag "mask flag" is a flag for controlling transmission of event data to the sound source 3, and is "1".
At the time of, it is instructed not to send the read event data to the sound source 3. The chord root "root" is a register that stores the root note of the chord specified by the key depression.
The chord type "type" is a register for storing the chord type designated by key depression (major, minor, etc.).

The manipulator process will be described in detail below with reference to the flowchart of FIG. FIG. 6 shows a part of processing for the operation of the start / stop switch 51, the fill-in switch 52, and the ending switch 53, among various operators.

Start / stop switch 5 in the operator
When it is determined that 1 has been pressed (step B1), it is determined whether or not the performance operation is being performed until then (step B2), and if the performance operation is being performed, the start / stop switch 51 is pressed this time. Means stop instruction,
In order to stop the performance, the interruption of the tempo clock is prohibited (step B4), and the musical tone currently being sounded is muted (step B5). On the other hand, if the performance operation is not in progress (step B2), the pressing of the start / stop switch 51 this time means a start instruction. Therefore, the accompaniment start preparation is made to start the performance (step B6), and the tempo clock interrupt occurs. Is permitted (step B7). Preparations for accompaniment start include clearing the tempo clock "clock", setting the accompaniment pattern to be played, and initializing flags. Then, the process returns to the main routine (step B13).

The fill-in switch 52 in the operator
When it is determined that is pressed (step B8), "1" is set to the status "status" (step B8).
9). As a result, the transition is made in the order of “current pattern → fill-in pattern → normal pattern”, and the performance of the normal pattern is continued thereafter. Further mask flag "
The mask flag "is set to" 0 "(step B10), and the process returns to the main routine.

The ending switch 5 in the operator
When it is determined that 3 is pressed (step B11), the status "status" is set to "2" and the process returns to the main routine. As a result, the sequence of "current pattern → fill-in pattern → maintaining the sound of the first musical tone of the normal pattern maintained for a predetermined time" is entered, and the performance ends.

FIG. 7 is a flow chart showing the key pressing process. In the apparatus of this embodiment, the keyboard 6 is divided into a melody area and a chord designation area for use. The entire area of the keyboard 6 is scanned (step C1), and if a key is pressed in the melody area (step C2), the sound source 3 is instructed to sound at the pitch corresponding to the key (step C3), and the melody area If there is a key release at (step C4), an instruction is issued to mute the tone at that pitch (step C5).

If there is a key depression in the chord designation area, the chord is detected from the key depression situation (step C6), and the root note and type of the detected chord are set to the chord root "ro".
Set to the register of ot "and code type" type ".
For example, if the detected code is A minor, the code root "root" = "A" and the code type "type" = "minor" are set.

Next, the tempo clock interrupt processing will be described with reference to FIGS.
This will be described with reference to the flowchart of FIG. The tempo clock interrupt process sets the tempo clock "clock" to 0 at the beginning of the accompaniment pattern and sets the tempo clock "clock" at the timing of the tempo clock generated by the tempo clock generator 2 to set the pattern length "le".
ngth "is sequentially incremented, and at each timing, it is determined whether or not there is event data in the accompaniment pattern, and at the timing when the event data exists, the event data processing (steps S1 to S8) is performed, and there is no event data. At the timing, instructions for the next performance pattern and processing for ending are performed (steps S9 to S30).

In the processes of steps S9 to S30, various registers are set so that the following processes are performed.
That is, during a normal performance, the normal pattern is selected again after the normal pattern and the performance is repeated. When the fill-in switch 52 is pressed, a "normal pattern → fill-in pattern → normal pattern" is displayed. Thus, the fill-in pattern is inserted during the performance of the normal pattern. Then, when the ending switch 53 is pressed, an ending process for ending the performance is performed by "maintaining the sound of the first tone of the normal pattern → fill-in pattern → normal pattern for a predetermined time". FIG. 1 is a diagram for explaining a manner of maintaining the pronunciation of a musical tone at the beginning of a normal pattern in this ending process. For example, when the normal pattern is as shown in FIG. 1) is played for a predetermined time (half a bar in this embodiment) as shown in FIG. 1B, and then the performance is ended. Hereinafter, processing for realizing these will be described with reference to the flowcharts of FIGS.

When an interrupt occurs at the timing of the clock generated by the tempo clock generator 2, the interrupt process is activated. First, the timing of the event data in the accompaniment data is referenced to see if there is event data to be read at the current tempo clock "clock" (step S
1).

If there is event data to be read, the type and part of the event data are checked and it is determined whether they are the following data, note on / off of chord part or note on / off of bass part (steps S2 and S4). ), If it is these event data, the note number of the event data is the code specified by the key depression in the above-mentioned key depression processing (chord root "root", chord type "type").
The pitch conversion is performed according to (steps S3 and S5).

The pitch conversion will be described below. Since the accompaniment data is stored in the accompaniment data memory 7 in the C major scale, the pitch of the accompaniment data is converted so as to match the chord designated by the key depression.

First, the alternation table 8 is used to match the code type "type". This is a table as shown in FIG. 11, where "C, C #, D ...
-A #, B "is to specify how much to shift each of the 12 tones in semitone units. This table is prepared for each code type "type". FIG. 11 shows major and minor. In the case of major, any note remains unchanged from the pitch of the accompaniment data, and in the case of minor, the pitch "E" of the accompaniment data is lowered by one semitone.

Next, the pitch of the accompaniment data is uniformly raised and lowered to match the chord root "root". For example, when the chord root "root" is "A", it is three semitones lower than "C", and therefore the pitch is uniformly lowered by three semitones. If the chord root "root" is "G", it is 7 semitones higher (or 5 semitones lower) than "C", so it is uniformly raised by 7 semitones (down by 5 semitones). By the above, pitch conversion can be performed.

Next, the state of the mask flag "mask flag" is judged (step S6). This mask flag "mask f
"lag" is for stopping (masking) the transmission of the event data to the sound source 3 when it is set to "1". Therefore, if it is not "1", the event converted from the accompaniment pattern and converted into a pitch The data is sent to the sound source 3 to instruct pronunciation and tone color change (step S7), and based on the note on / off and tone color change instructions sent to the sound source 3, which part of which sound source is currently in the sound source 3. Information on whether or not high is pronounced is stored in the pronunciation status map (step S
8). On the other hand, when the mask flag "mask flag" = "1", the event data is not sent to the sound source 3, so that the accompaniment pattern cannot be played ahead of the current position.

On the other hand, in step S1, the tempo clock "
If there is no event data corresponding to "clock",
Add one tempo clock "clock" (step S
9), it is determined whether or not the status "status" is "4" (step S10). The status "st
"atus" = 4 means that the process of maintaining the pronunciation of the first tone of the last normal pattern in the ending process for a predetermined time is performed. Therefore, the status "status"
If not equal to 4, then it is determined whether the tempo clock "clock" has reached the pattern length "length" (step S1).
1). Tempo clock "clock" is pattern length "length"
If it has not reached, the accompaniment pattern has not ended yet (that is, the middle of the accompaniment pattern is still playing), so the interrupt processing is ended and the process returns to the main routine (step S11).

The tempo clock is determined by the judgment in step S11.
When "clock" reaches the pattern length "length", the performance of the accompaniment pattern has ended, and therefore the processing for setting the next accompaniment pattern is performed (steps S17 to S30). status "
The next accompaniment pattern (normal pattern or fill-in pattern) is selected according to the value of, and if ending is instructed, a series of control of ending processing is performed.

First, it is determined whether or not the status "status" = 0 (step S17). Status "stat
Since "us" = 0 means a normal accompaniment performance mode, when the status is "status" = 0, a normal pattern is selected as the accompaniment pattern to be read next from the accompaniment data memory 7 (S18). The read address of the accompaniment pattern to be read is set (step S28), and the pattern length written in the accompaniment pattern is set to the pattern length "le".
Write to the ngth "register (step S29) and clear the tempo clock" clock "to zero (step S3).
0).

If the status "status" is not 0, it is further determined whether the status "status" is 1. Since the status "status" = 1 means that the fill-in switch 52 is pressed to instruct the performance of the fill-in pattern, when the status "status" = 1 is determined, the fill-in pattern is selected as the accompaniment pattern to be read next. Select (step S20), and then status "statu
s "= 0 is set (step S21), and the above steps S28 to S30 are performed, whereby the performance of" normal pattern → fill-in pattern → normal pattern "is performed.

If the status "status" is neither 0 nor 1, it is next determined whether the status "status" is 2 (step S22). The status "status" = 2 means that the ending switch 53 is pressed and the ending process is instructed. In this case, the performance is ended by "maintaining the sound of the first musical tone of the normal pattern → fill-in pattern → normal pattern for a predetermined time", and the following processing for this is performed.

First, the fill-in pattern is selected as the accompaniment pattern to be read next (step S23), and the status
"status" = 3 is set (step S24). Then, when the fill-in pattern is completed, the status "status" = 3 is determined in step S25, and at that time, the normal pattern is selected as the accompaniment pattern to be read next (step S26), and the status "st" is set.
Atus "= 4 is set (step S27).

As a result, when the performance of the normal pattern is started, the status "statu
It is determined that s "= 4 (a state in which the tone generation maintaining process is performed after shifting to the normal pattern), and thereafter, the process for time extension of the musical tone at the beginning of the normal pattern (steps S12 to S).
16) is performed.

First, it is determined whether or not the tempo clock "clock" has reached a predetermined number (step S12).
Set 1 to the mask flag "mask flag" (step S1
3) After that, the read event data is not sent to the sound source 3. That is, the musical tone for which note-on has been sent (sounding) up to this point is set as the leading musical tone of the normal pattern, and the musical tone is maintained thereafter. In step S12, a comparison is made with a predetermined number of tempo clocks "clock". This may be the case where the musical tone at the beginning of the musical tone at the beginning deviates from the timing of tempo clock "clock" = 0 depending on the data of the performance pattern. The event data (note-on) may exist, so the tempo clock "clock" comparison has a width in order to generate such a musical sound.

Next, it is determined whether or not the count of the tempo clock "clock" has reached a predetermined time (half a bar is set in this flowchart) for maintaining the pronunciation at the ending (step S14), and the count has been reached. If not, the interrupt process is repeated for each tempo clock until it is reached, and when it is reached, the tempo clock interrupt is prohibited (step S1).
5) Stop playing further (stop playing). Next, a note-off is sent to the sound source 3 with reference to the pronunciation situation map, and the musical tone being sounded by the sound source 3 is muted (step S16).

In carrying out the present invention, various modifications as described below are possible. (1) The process may be shifted to the ending (the process of maintaining the sound generation and ending) without passing through the fill-in. In the embodiment, when the ending switch 53 is pressed, the fill-in pattern is once played and then the normal pattern is entered, so that the first musical tone of the normal pattern is continuously generated. However, the present invention is not limited to this, and when the ending switch 53 is pressed, it is possible to shift to the normal pattern without passing through the fill-in pattern and continuously generate the musical tone at the beginning.

(2) The ending process may be performed for only some of the parts. In the embodiment, the ending process is performed without distinguishing the chord part, the bass part, and the drum part in the accompaniment data, but the present invention can be applied to each part. For example, the drum part plays the ending pattern as in the conventional case, while the present invention is applied to the chord part and the bass part in parallel, and the ending pattern is obtained from the fill-in pattern and the normal pattern. You can also do it.

(3) The continuous sound to be generated may be rounded into the constituent sounds of the designated chord. In the embodiment, in the ending process, the leading musical tone of the normal pattern is sounded for a predetermined time, but when the leading musical tone is a sound other than the constituent sound of the designated chord, I can't get the feeling that the performance ends even if I keep on playing. For example, when the normal pattern is as shown in FIG. 12 and "C" is specified as the chord, the leading musical sound "B" is continuously pronounced. The constituent sounds of "C,
It is not included in "E, G", and a feeling of end cannot be obtained even if the pronunciation is continued. In such a case, it is advisable to convert the musical sound for which the pronunciation is to be sustained so as to be one of the constituent sounds of the chord. In the above example, the musical sound "B" may be converted into the nearest chord constituent sound "C".

As another method, the normal pattern may be played from the beginning, and when the chord constituent tone is found, the musical tone may be continued. Figure 1 above
In the example of 2, the musical tone "B" is pronounced and then the musical tone "C" is pronounced. However, since this is included in the constituent tones of the code "C", if the musical tone "C" is continuously pronounced. Good. As described above, the ending performance can be performed with a musical sound suitable for obtaining a feeling of end.

(4) The sounding time of a continuous sound, which is determined according to the performance tempo, may be corrected. As in the embodiment, the duration of the musical tone can be set to half a bar, or set to 1 beat, but when set to 1 beat, the time changes depending on the performance tempo. To do. For example, if the tempo is half, the time for one beat is doubled, and the sounding time of a continuous sound is inversely proportional to the performance tempo. However, when the tempo is slowed down, it is more likely that the continuous tone is sounded for a longer time, so in this case, the sound is generated for a time longer than one beat.

[0057]

As described above, according to the present invention, it is not necessary to prepare a predetermined ending pattern in advance for ending the performance, so that the memory is not consumed for storing the ending pattern. In addition, it is possible to generate a musical sound that reminds the end of the performance with a simple configuration, and it is possible to realize the end of the performance with a preferable ending feeling.

[Brief description of drawings]

FIG. 1 is a diagram illustrating an outline of the present invention.

FIG. 2 is a diagram showing an automatic performance device as one embodiment of the present invention.

FIG. 3 is a diagram showing a configuration example of accompaniment data in the device of the embodiment.

FIG. 4 is a diagram showing details of an operator group in the apparatus of the embodiment.

FIG. 5 is a flowchart showing a main routine in the embodiment apparatus.

FIG. 6 is a flowchart showing a manipulator processing routine in the embodiment apparatus.

FIG. 7 is a flowchart showing a key depression processing routine in the embodiment apparatus.

FIG. 8 is a flowchart (1/3) showing an interrupt processing routine in the embodiment apparatus.

FIG. 9 is a flowchart (2/3) showing an interrupt processing routine in the embodiment apparatus.

FIG. 10 is a flowchart (3/3) showing an interrupt processing routine in the embodiment apparatus.

FIG. 11 is a diagram showing an example of an alternation table.

FIG. 12 is a diagram showing an example of a normal pattern in a modified example.

[Explanation of symbols]

1 CPU 2 Tempo clock generator 3 sound sources 4 Display 5 operators 6 keys 7 Accompaniment data memory 8 Alteration table 9 work areas 10 program memory 11 amplifier 12 speakers

─────────────────────────────────────────────────── ─── Continuation of the front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) G10H 1/00-7/12

Claims (2)

(57) [Claims] 1. A storage means for storing a plurality of performance patterns different from an ending pattern, the performance patterns including pitch information in association with a style, and a performance pattern corresponding to the style is read from the storage means and automatically stored. After the performance means for performing the performance, the performance end instruction means for instructing the performance end, and the performance end instruction by the performance end instruction means, the head part of the predetermined performance pattern corresponding to the style is read from the storage means. , The musical tone stored in the head portion is sounded at a pitch according to the pitch information included in the predetermined performance pattern, and the performance is ended after the sound has been sounded for a predetermined time not depending on the predetermined performance pattern. And an automatic performance device having a performance control means for controlling the performance means. 2. A storage means for storing a plurality of performance patterns different from the ending pattern, the performance patterns including pitch information corresponding to the style, a chord designating means for designating a chord, and a style from the storage means. A performance means for reading out a performance pattern corresponding to the above, automatically converting the pitch information according to the designated chord, a performance end instruction means for instructing the end of the performance, and a performance end instruction means for performing the performance. After the end instruction, the head portion of a predetermined performance pattern corresponding to the style is read from the storage means, and the musical tone stored at the head portion is converted into pitch information included in the predetermined performance pattern. At the pitch converted according to the chord designated by the chord designating means, the performance is ended after the sound is produced for a predetermined time not depending on the predetermined performance pattern. An automatic performance device having a performance control means for controlling the performance means.