JP2733161B2 - Training device with automatic performance piano - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a training apparatus using an automatic performance piano, and more particularly, to a comparison between a model performance and a user's own performance. The present invention relates to a training apparatus that can easily perform self-study by notifying a performance correction instruction for eliminating a mode of disagreement.

[0002]

2. Description of the Related Art Conventionally, as disclosed in Japanese Patent Application Laid-Open No. 60-111296, at least one of the musical instruments is provided with a comparing means for comparing a plurality of pieces of musical tone information stored in a storage means. Electronic musical instruments that can be evaluated for performance are known. For example, musical tone information (scale data and pitch data) of a performance by a skilled player is stored in advance,
Compare with your own music information.

[0003] A scale comparator for comparing scale data and a pitch comparator for comparing pitch data are provided. The scale comparator compares scale data of a performance by a skilled performer and scale data of a user's own performance. Output is made when there is a mismatch, and the pitch comparator is configured to output when the pitch data of the performance by the skilled player and the pitch data of the own performance do not match. Then, by counting the number of unmatched outputs, subtracting it from the full score, and scoring the performance, it is possible to compare the superiority and enjoy the performance.

[0004]

However, in the prior art, the two performances are simply compared, and how much the musical information such as the scale and the pitch differs (some differences are different) or Even if the address at which the mismatched output is stored is stored, it is only known which part is different. Therefore, even if you compare the example performance with your own performance, it is not immediately clear how to fix the part that was different from the example performance. One is not effective.

Accordingly, the present invention solves the above-mentioned problems, and finds out the bad part of one's own performance, as well as how to fix it to get closer to a model performance, and to effectively carry out self-study. It is an object of the present invention to provide a training device using an automatic performance piano that can be performed easily.

[0006]

To achieve Means for Solving the Problems] Such object, an invention according to claim 1, Example Shimesuru so in Figure 1, as performance information along the flow of playing the key depression data, and key-off data Creates key press data and key release data from the stored performance information storage means M1, key press and key release timing detected based on the key operation position, and key hitting strength information calculated based on the key press speed. The performance information storage means M2 is stored in the performance information storage means M1 along the flow of the performance based on the key press and key release timing, and is stored in the performance information storage means M1 along the flow of the performance. The individual key press data and the key release data in the performed performance information are sequentially output, and the solenoid is driven by the driving power determined based on the keying strength information in the output key press data, and the key release data is corresponded. And Solenoi And playing reproduction means M3 for reproducing a performance by stopping the drive, an automatic player piano comprising a reference play information storage means stores a predetermined key depression data and key-off data as reference play information M4
And a mismatch mode determining unit M5 that compares the performance information of the comparison target stored in the performance information storage unit with the model performance information with respect to a predetermined check point, and determines a mode of mismatch between the two pieces of performance information. The correction instruction storage means M6, which stores in advance a performance correction instruction for eliminating the mismatch mode in accordance with the mismatch mode, and the presence of the mismatch when there is a mismatch between the two pieces of performance information. At the check point to be performed, the correction instruction notifying means M7 for notifying the corresponding performance correction instruction and the performance reproducing means M3.
The performance of the comparison target is reproduced, and the performance of the
Updating the corresponding performance correction instruction according to the performance timing
Notification control means to be notified by the correction instruction notification means M7
And M8, the training device by the automatic player piano, characterized in that it includes a as its gist.

[0007] As the above-mentioned predetermined check point,
"P (piano)", "f"
(Forte) "," tenuto "," legato ", and the like. Further, the degree of the mismatch may be divided into a plurality of stages, and the performance correction instruction may be changed for each of them and stored in the correction instruction storage means.

[0008] Examples of the correction instruction informing means, may be appealed to the visually displayed on the image, but it may also be to appeal to the hearing to notify by voice.

[0009]

According to the training apparatus using the automatic performance piano of the present invention having the above-described configuration, for example, when a student plays the automatic performance piano, the performance information storage means M2 causes the key press data and the key release data to flow. Is stored in the performance information storage means M1 as the performance information along. Then, the non-coincidence mode determination means M5 compares the performance information of the student stored in the performance information storage means M1 with the model performance information stored in the model performance information storage means M4 for a predetermined check point. Then, the mode of mismatch between the two pieces of performance information is determined.

If there is a mismatch between the two pieces of performance information,
Correction instruction notification unit M7 is, in the check points present the mismatch be informed by reading the playing correction instruction corresponding than correction instructions storing means M6 stored in advance to play correction instruction for eliminating the aspects of the mismatch In
In this training device, the notification control means M8
Reproduces the performance of himself (student) by the performance reproducing means M3
To match the performance timing of the mismatched part
The corresponding performance correction instruction is notified by the correction instruction notifying means M7.
Let me know.

The determination of the above-mentioned mismatch mode is performed by comparing the pitches of the sample performances, and comparing the volume, the length of the notes, and the like according to the instruction content of the performance instruction symbol. For example, in a portion having a dynamic sign such as p (piano) or f (forte), the judgment is made based on the volume, and “tenuto”,
For "legato" etc., it is determined based on the length of the sound.

As described above, if a performance correction instruction for a mismatched portion is displayed on an image and appealed visually, or notified by voice and appealed to the auditory sense, the student can recognize the bad part of his or her performance, and You can see how you can approach the model performance, so that self-study can be done effectively.

In the present training apparatus, it is possible to notify the corresponding performance correction instruction in accordance with the performance timing of the mismatched portion while reproducing the performance of the own (student).
Therefore, the student can listen to the performance correction instruction while objectively listening to his own performance in the same state as when he actually played it. Therefore, the mode and degree of the disagreement can be grasped more intuitively, and self-study self-study can be performed more effectively. Also,
For example After reproduce their performances by playing reproducer M3, if caused to reproduce the reference play by play reproduction means M3, you can hear more carefully playing model in a portion of the performance correction instruction, It becomes easy to intuitively learn the loudness and length of the sound of the model performance.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A training apparatus for an automatic performance piano according to one embodiment of the present invention will be described. In the automatic performance piano 1 as an embodiment, as shown in FIG. 2, a stepped shutter 5 attached to the lower surface of a key 3 blocks the optical paths of two key sensors 7, 8 each including a light emitting element and a light receiving element. The control unit 10 detects a time and a time difference, and creates performance information in which a keyboard number, a key press timing, a key release timing, and a keystroke intensity are combined.

As shown in FIG. 3, the control unit 10
CPU 11, ROM 12, RAM 13, backup R
A logic operation circuit including an AM 14, a clock 15, etc.
The key sensors 7 and 8 are connected via an input / output port 16. Also, a floppy disk driver 23 for driving a floppy disk 21 for recording automatic performance data or for recording new automatic performance data, a solenoid 25 for key driving during automatic performance, and various operations such as recording and playback. An operation panel 27 for instructing the display, a display 29 for display, a sound generator 31 described later, and the like are also connected via the input / output port 16.

The automatic performance piano 1 of the present embodiment having the above-described hardware configuration can store the performance of the player in the procedure as shown in FIG. This processing is performed on the operation panel 27.
Is started when the recording mode is selected in. In the embodiment, recording by a so-called event method and reproduction to be described later are performed.

First, detection signals from passage detection sensors 7 and 8 (only one is shown in FIG. 2 but arranged for all 88 keys) reflecting the key press / release operation of the player are shown. It is determined whether or not it has changed (Step 110, hereinafter simply referred to as S11
It is called 0. The same applies hereinafter. If there is a change, the ON event occurrence time ton is detected from the time when the detection value of the left key sensor 7 changes from ON to OFF, and the detection value of the right key sensor 8 changes from OFF to ON, depending on the manner of the change. From the time when the off-event occurrence time is reached (S1
20).

Subsequently, it is determined whether or not the current change in the detection signals of the key sensors 7 and 8 is a change due to key depression (S13).
0). If the on-event occurrence time point ton is grasped in S120, it is determined that the key is pressed, and the time when the detection value of the right key sensor 8 in the current key sensor change is changed from on to off, and the detection of the left key sensor 7 The time difference Δt from the time when the value changes from on to off is calculated (S14).
0).

Subsequently, keying strength data Pon is calculated based on the time difference Δt (S150). Here, K is a coefficient for calculation, and is set in advance based on the positional relationship between the steps of the shutter 5 and the like. Next, the keying strength data P
The sound generation delay time TDrec-on during recording is calculated based on "on" (S160). This indicates that the key 3 has moved to a specific position in the sounding operation at the time of occurrence of the ON event, but the actual sounding is performed by mechanically driving the hammer system, and the moving speed of the key 3 This causes a difference from the time when the on-event occurs.

The keying strength data Pon is stored in the key number KNo.
Then, key press data Don is created in combination with the on-event occurrence time point ton (S170). On the other hand, if it is determined in S130 that the key is not depressed, that is, the current key sensor change is a change due to key release, the off-event occurrence time tof
Silence delay time TDre from f to the actual silence timing
c-off is set (S180).

Then, similarly to the key press data Don, the key number KNo. And the keystroke strength data Pon are combined with the key off event occurrence time toff to create key release data Doff (S170). The keystroke strength Pon in the key release data Doff is usually "0". The sounding delay time TDrec-on is added to the key press data Don thus calculated to produce sounding timing information Don + TDrec, and the sound release timing Don is added to the key release data Doff to add the sound stopping delay time TDrec-off. Information Doff + TDrec, and these are R
The data is temporarily stored in a free channel in the dedicated assigner in the AM 13 (S190). In the embodiment, one assigner is constituted by 64 channels.

Here, the sound generation delay time during recording TDrec-on
Is the time from the occurrence of the ON event detected by the player's key depression to the actual sound generation time. Therefore, FIG.
As shown in (1), the data tends to be longer as the keying strength Pon is smaller, and is within a range of several msec to 100 msec. In the present embodiment, the sounding delay time TDrec-on during recording is measured in advance by changing the keying strength Pon, and the result is mapped and written in the ROM 12, and S14 is executed.
The value is specified by referring to the keying strength Pon calculated at 0 and using an interpolation method or the like.

On the other hand, the recording stop sound delay time TDrec-off is fixed to 35 msec in this embodiment. This is a representative value of the return speed based on the mechanical characteristics of the key 3 and the action unit at the time of key release. As a result of the processing of S190, information having delay time data TDrec in the first byte, key number data KNo. In the second byte, and keying strength data Pon in the third byte, as shown in FIG. As hexadecimal notation. This figure shows only the key press data Don, but the same applies to the key release data Doff.

Next, it is determined whether or not a predetermined short time ΔTD (5 msec in the embodiment) has elapsed (S20).
0). If “NO” is determined, the process returns to S110. If “NO” is determined in S110, S1
It is configured to directly go to S200 after passing S20 to S190.

If "YES" is determined in the process of S108a, the minute time ΔTD is subtracted from the delay time data TDrec of the information temporarily stored in the channel to be processed first in the recording assigner. (S
210), the delay time data TDrec after the subtraction is 0 ms
It is determined whether or not c has been reached (S220).

If "YES" is determined, the information is transferred to the floppy disk 21 as performance information (S230), and S190, S22 for all channels.
It is determined whether or not the process of 0 has been executed (S240). If there is any remaining channel, a predetermined value ΔTD is subtracted from the delay time data TDrec (S210), and the processes of S220 and thereafter are repeatedly executed. If “NO” is determined in S220, the process proceeds to S240 by passing S230.

As described above, by the recording process of the present embodiment, the performance information is recorded in relation to the actual sound generation timing and the actual sound stop timing. That is, the on-event occurrence ton
In addition to the fact that the off-event occurrence toff is not merely caught by the mechanical operation relationship, it is possible to use the information obtained by adding the correction based on the reality and replacing the information with the relationship between the actual sounding timing and the sound stopping timing.

Next, the recorded performance information of the user (hereinafter referred to as "self performance information") is compared with model performance information (hereinafter referred to as "model performance information") at predetermined check points. The process of notifying the performance correction instruction will be described. The model performance information may be read from the floppy disk 21a in which the model performance information is stored in advance, or may be used by a skilled person such as a piano teacher playing the automatic performance piano 1 and performing the above-described recording processing. You may use what was recorded.

The points to be checked include whether the melody matches, whether the length of a note (including a rest) is correct, whether the performance indicated by the performance instruction symbol is performed, and the like. Hereinafter, as examples, comparison / report control processing relating to melody and comparison / report control processing relating to performance instruction symbols will be described.
The two processes will be described.

First, the comparison / report control processing relating to the melody will be described with reference to the flowchart of FIG. This process is started when the melody comparison notification mode is selected on the operation panel 27. First, key number data KNo. Is read from the self performance information and the model performance information to be compared (S310). Then, the corresponding key number data KNo. Are sequentially compared with each other from the beginning of the performance along the flow of the performance (S320), and it is determined whether or not the key numbers match (S330). If they match, it is determined whether the comparison of the last data has been completed (S3).
50) If not finished, return to S320 and repeat the following processing. If the two key numbers do not match in S330, the mismatched key number data KNo. Is stored in the RAM 13 (S340).
The process moves to S350.

When the last data has been compared (S350: YES), it is determined whether or not the mismatched key number data KNo. Is stored in the RAM 13 (S360). If the mismatched data exists (S360: YES), the key number data KNo. For several measures including the note corresponding to the mismatched key number data KNo. Is read from the model performance information. In addition, the musical score information is displayed on the display 29, and a non-matching note corresponding to the key number data KNo. On the self-performance information side is blinked on the musical score information (S370). on the other hand,
If there is no unmatched data (S360: N
O), "No mismatch" is displayed on the display 29.

If there are many unmatched portions and cannot be displayed on the display 29 at one time, they may be scrolled in order and viewed. Further, the range displayed on the display 29 as the musical score information is several bars including the mismatched note in the above embodiment, but this is to make it easier to grasp the flow of the performance before and after the mismatched note.
Therefore, for example, the attached time signature (two quarter time,
6/8) or the like.

As described above, the student can see which part of the melody differs between the model performance and the self-performance by looking at the musical score information on the display 29, and can learn the correct melody. And this performance training
To display the correct instruction on the display 29,
Read the self-play performed from the floppy disk 21 and
To match the timing of the mismatched notes
To flash the corresponding mismatched note over the score information.
You. It should be noted that, regarding the mode of this notification, the sound generation device 3
The information may be notified from 1 and appeal to the user's hearing. This
Students can actually perform their own performance
Instructions to correct performance while listening objectively in the same state as when playing
You can see and hear. For example, playing correction
Even if the instructions point out that the notes do not match,
It is difficult to understand whether such instructions were given to the performance
No. However, in this embodiment, one's own performance is regarded as one's own eyes.
Listen to the instructions while listening objectively with
Can be more sensitive to the mode and degree of discrepancy
Self-study can be done more effectively.
Also, reproduce your own performance while notifying the performance correction instruction
After that, if you reproduce the sample performance,
Listen more carefully to the example performances where indicated
And hand to the bad part of your performance that you just heard
Easy to learn more about the loudness and length of the performance
It becomes. In this way, listening, which is essential for music practice,
Since learning through the sense can be done by oneself, training equipment and
It is very effective. Next, a comparison / notification control process relating to performance instruction symbols will be described with reference to flowcharts of FIGS. In this embodiment, a performance instruction symbol is stored in advance in the model performance information. The performance instruction symbols are classified into three groups of a volume group, a tone length group, and a performance speed group in advance, and which performance instruction symbol belongs to which group is stored.

The volume groups are “p (piano)”, “f”
(Forte), etc., related to volume, and the duration group is “Staccato”, “Tenuto”
Etc., relating to the length of the note. The performance speed group relates to the performance speed in general. In addition to “rit. (Ritardando)”, a speed symbol for designating the performance speed,
Also includes speed slogans such as "Andante" and "Allegro".
For these performance instruction symbols, numbers are set so that the groups to which they belong and the instruction content can be distinguished, and the numbers are given to the key number data KNo. Of the corresponding notes so as to correspond to the actual musical score.

This processing is started when the performance instruction symbol comparison notification mode is selected on the operation panel 27. First,
Instruction symbols are sequentially read from the model performance information (S41).
0), if it belongs to the volume group (S420: YE
S), and executes a volume comparison processing routine to be described later (S4).
30) If it belongs to the note length group (S44)
0: YES), a sound volume comparison processing routine to be described later is executed (S450), and it is not a sound volume group (S420: N).
O), if it is not a note length group (S440: N
O), a performance speed comparison processing routine is performed (S460).

After performing the comparison process according to the group to which each symbol belongs in this way, it is determined whether the comparison process for the last performance instruction symbol has been completed (S47).
0) If not finished, return to S410 and repeat the following processing. If the comparison process for the last performance instruction symbol has been completed (S470: YES), the process proceeds to S430.
As a result of the processing in the processing routine of S450 and S460, it is determined whether or not a performance correction instruction is stored in the RAM 13 (S480). Then, when the performance correction instruction is present (S480: YES), the performance correction instruction is displayed on the display 29 (S490), and when there is no performance correction instruction (S480: N).
O), "No mismatch" is displayed on the display 29, and the process is temporarily terminated. In the present embodiment, as a method of displaying the musical score on the display 29 in S490, the musical score is sequentially displayed on the display 29 based on the model performance information, and when the corresponding musical symbol is present, the musical score does not overlap with the musical score. In this way, words of a performance correction instruction described later are displayed.

Next, each comparison processing routine (S430, S430)
450, S460) will be described in detail in order. The flowchart of FIG. 9 shows the volume comparison processing routine of S430.
When the main volume comparison process is started, first, a note is designated from the self performance information and the model performance information in accordance with the instruction content of the performance instruction symbol (S431). This note specification is key number KN
This is done by specifying o. Then, the keying strength Pon corresponding to the designated key number KNo. Is read out from both performance information (S432).

Then, the key strength Pon of the model performance information is subtracted from the key strength Pon of the self-performance information to calculate a strength difference PS (S433). The ranking is performed based on the calculated intensity difference PS and the content of the performance instruction (S434), and a performance correction instruction corresponding to the content of the instruction and the rank is read and temporarily stored (S435). In advance, the intensity difference P
A rank classification corresponding to S and a performance correction instruction corresponding to the rank are stored as a table.

FIGS. 11 (a) and 11 (b) show ranking and correction instruction tables for "pp (pianissimo)" and "ff (fortessimo)" among the dynamic symbols. FIG.
The table in the case of pp (Pianissimo) in (a) will be described in detail. In the case of pp (Pianissimo), the predetermined value α is used, and in the case of α ≧ PS ≧ −α, the keying strength Pon of both pieces of performance information is regarded as an allowable range that can be considered to be the same. As the value increases to B2, the determination upper limit value is increased by a predetermined value α. When the intensity difference PS is larger than 3α, it corresponds to rank B3. On the other hand, when the intensity difference PS is smaller than -α, it corresponds to rank S1.

A performance correction instruction is set according to each rank. For example, in the case of rank S1, the keystroke strength Pon of the self-performance information is slightly weaker than the keystroke strength Pon of the sample performance information.
In order to get closer to, a correction instruction of "just a little stronger" is set.

On the other hand, in the case of the rank B1, it is "slightly weak", in the case of the rank B2, it is "slightly weaker", and in the case of the rank B3, it is quite strong that the user does not notice the instruction p (piano) at all. Since the keying strength is Pon, each correction instruction of “too strong, weaker” is set. In the case of rank N, there is no correction instruction because the keying strengths Pon of both pieces of performance information are within an allowable range that can be regarded as coincident.

In the case of ff (fortessimo) in FIG. 11B, the ranks B1, N, S1 to S3 are determined by using a predetermined value β.
The performance correction instructions corresponding to each rank are set. In the case of pp (pianissimo) described above, the keystroke strength Pon of the self-performance information is set to three levels (B1 to B3) that are larger than those in the model performance, while ff
In the case of (Fortissimo), three steps (S1 to S1
3) It is set. This is the pp
This is because the player tends to play strongly despite the performance instruction of (Pianissimo), while he tends to play weakly despite the performance instruction of ff (Fortissimo). In the case of pp (pianissimo), a small volume difference becomes a problem as compared with the case of ff (fortissimo), so the above-mentioned predetermined value α is often set to a value smaller than the predetermined value β.

For other dynamic symbols, for example,
mp (meso piano) and mf (meso forte) are two steps (B1, B2) for the larger one and two steps (S for the smaller one).
1, S2) may be appropriately set according to the content of the instruction. If “crescendo” or “decrescendo” is added, the keying strength data Pon of one or more notes in each bar is read out for several bars after the marked portion. And compare them. Then, a performance correction instruction such as "slightly stronger (weaker)" or the like may be temporarily stored in the same manner as described above.

After temporarily storing the performance correction instruction in step S435, the present volume comparison processing routine is temporarily terminated, and the process returns to step S435 in FIG.
The process moves to 470. Next, the sound length comparison processing routine of S450 will be described with reference to the flowchart of FIG. When the actual sound length comparison process is started, first, a note is designated from the self-performance information and the model performance information in accordance with the instruction content of the performance instruction symbol (S451). This note designation is performed by designating the key number KNo. Then, based on both performance information, the sounding timing information Don corresponding to the designated key number KNo.
+ TDrec and sound stop timing information Doff + TDrec are read out (S452).

Then, the above tone generation timing information Don + T
The note length L1 on the self performance information side and the note length L2 on the model performance information side are calculated based on Drec and the sound stop timing information Doff + TDrec (S453). continue,
The note length L2 on the sample performance information side is subtracted from the note length L1 on the self performance information side to calculate a note length difference LS (S45).
4). A ranking is performed based on the calculated length difference LS and the content of the performance instruction (S455), and a performance correction instruction corresponding to the content of the instruction and the rank is read out and temporarily stored (S45).
6). Also in this case, the pitch difference LS is previously set for each performance instruction symbol.
Are stored as a table, and a performance correction instruction corresponding to the rank is stored as a table.

FIGS. 12 (a) and 12 (b) show ranking and correction instruction tables for "Staccato" and "Tenuto". “Staccato” is an instruction to perform a short cut for each note, but in actual performance, it tends to be played long regardless of the “Staccato” performance instruction. Therefore, in the present embodiment, it is assumed that there is no case where the length is too short.
It is assumed that the note lengths L1 and L2 are in the allowable range that can be regarded as coincident, and the rank is N. Those with a pitch difference LS of 2γ or less are ranked B1, and those with a pitch difference LS greater than 2γ are rank B2.

In the case of rank B1, since the note length L1 on the self performance information side is slightly longer than the note length L2 on the model performance information side, a correction instruction of "slightly shorter" is set. ing. In the case of rank B2,
Since the note length L1 on the self-performance information side is such that the staccato instruction is not noticed at all, a correction instruction of "too long, shorter" is set.

Similarly, the tenuto in FIG. 12 (b) is an instruction to keep the length of the note sufficiently, but in an actual performance, it is played short despite the performance instruction of "tenuto". Tends to. Therefore, in this embodiment, it is assumed that there is no case where the length is too long, and when the note length difference LS ≧ −ε is used by using the predetermined value ε, the two note lengths L1 and L2 are regarded as allowable ranges that can be regarded as coincident. And In addition, those having a tone length difference LS of up to -2.epsilon.

In the case of rank S1, since the note length L1 on the self-performance information side is slightly shorter than the note length L2 on the model performance information side, a correction instruction "a little longer" is set. ing. In the case of rank S2,
Since the note length L1 on the self-performance information side is short so as not to notice the instruction of "tenuto" at all, a correction instruction of "too short, longer" is set.

After temporarily storing the performance correction instruction in S456, the actual tone length comparison processing routine is temporarily terminated, and
The process moves to 470. Although the performance is correct from the viewpoint of the whole music, it is possible to judge that most of the music is incorrect if the tempo itself is mistakenly performed and a simple performance comparison is made. There is also. Therefore, it is preferable to first compare the playing times of the music pieces of the self-performance and the model performance, correct the note length L1 and the like on the self-performance information side based on the ratio, and then compare the tone.

Next, the performance speed comparison processing routine will be briefly described. For example, in the case of "ritardand (gradually gradual)", one or more notes in each measure are specified for the measure indicated by the symbol. Then, the length between the designated notes is calculated, and the difference between the note length on the self performance information side and the note length on the sample performance information side is obtained. Based on this inter-note length, similar to the above-mentioned note length comparison processing, it is only necessary to temporarily store a performance correction instruction corresponding to the rank, for example, an instruction such as "slightly slower".

If the specified range of the speed is wider, such as a speed symbol or a speed slogan, a plurality of notes are appropriately specified from the specified range (for example, the whole music or a certain movement unit) and specified. Calculate the length between the performed notes. Then, in the same manner as in the case of the above “ritardando”, ranking is performed based on the difference between the note length on the self-performance information side and the note length on the sample performance information side, and the corresponding performance correction instruction is temporarily stored. It is good.

In calculating the length between notes,
Sounding timing information Don + TDrec of the designated note,
Alternatively, it may be calculated using any one of the sound stop timing information Doff + TDrec. As described above, the model performance information and the self-performance information are compared in accordance with the instruction contents of the various performance instruction symbols. If there is a mismatch, the score is displayed on the display 29 in order.
In a portion where the corresponding performance instruction symbol exists, a word of a performance correction instruction for making a correct performance is displayed. For this reason, the student can understand which part of the performance instruction symbol is different between the sample performance and the self-performance, and can also see how much should be corrected by looking at the performance correction instruction. In other words, words having almost the same content as the correction instruction spoken by the teacher in a piano classroom or the like are displayed, so that self-study can be performed very effectively. The above-described rank division and the correction instruction corresponding to the rank can be appropriately increased or decreased.

Then, the performance correction instruction is displayed on the display.
In the case of displaying on the display 29 , the corresponding performance correction instruction is displayed on the display 29 in accordance with the performance timing of the mismatched portion while reading and reproducing the recorded self performance from the floppy disk 21 . Note that this correction
Regarding the mode of notification of the instruction, the voice generation device 31
It may be something that informs and appeals to the user's hearing.

With this configuration, the student can listen to the performance correction instruction while objectively listening to his own performance in the same state as when he or she actually played. For example, even if there is an instruction such as "stronger" or "longer" in the performance correction instruction, it is difficult to understand what kind of performance the user has issued such an instruction. However, in this embodiment, since the user can listen to the instruction while objectively listening to his own performance with his own ear, he can more intuitively grasp the aspect and degree of the discrepancy, and can perform self-study more effectively. .

If the user reproduces his / her own performance while notifying the performance correction instruction, and then reproduces the model performance, the user can listen more carefully to the example performance at the part where the performance correction instruction is given. This makes it easier for the user to more intuitively learn the loudness, length, and the like of the sample performance for the bad part of his performance that he heard immediately before. In this way, learning through hearing, which is indispensable in music practice, can be performed by oneself, so that it is very effective as a training device.

Although one embodiment of the present invention has been described above, the present invention is not limited to this embodiment, and various embodiments can be adopted without departing from the gist of the present invention. For example, it is not always necessary to execute all of the above-described various comparison and correction instruction notification processes, and a setting switch may be provided so that the user can freely select from among them according to his / her ability and learning stage. For example, if a person who has mastered playing the melody for the first time puts emphasis on performance instruction symbols in order to give an expression to the performance, the above-described comparison and correction instruction notification processing regarding the performance instruction symbols will be performed. May be executed. Also, add a printer,
The musical score and the performance correction instruction displayed on the display 29 may be printed out.

[0058]

As described above, according to the present invention, the performance information inputted by the student playing the automatic performance piano is stored, and the student performance information and the sample performance information are compared with respect to a predetermined check point. Then, the mode of mismatch between the two pieces of performance information is determined. If there is a mismatch between the two pieces of performance information,
At the checkpoint where the mismatch exists, a performance correction instruction for eliminating the mismatch mode can be notified. And while reproducing his own performance,
Appropriate performance correction corresponding to the performance timing of the critical part
Instructions can be given so students can play
Perform while listening objectively in the same state as when you actually played
You can hear the performance correction instructions. Therefore, the state of disagreement
Thus, the degree can be grasped more intuitively and self-study can be effectively performed.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating the configuration of a training device using an automatic performance piano according to the present invention.

FIG. 2 is a schematic configuration diagram showing a configuration for calculating a keystroke strength in the automatic performance piano of the embodiment.

FIG. 3 is a schematic configuration diagram showing a connection relationship between a control unit and each unit in the automatic performance piano of the embodiment.

FIG. 4 is a flowchart of a recording process in the embodiment.

FIG. 5 is an explanatory diagram showing a tendency of a map referred to in a recording process.

FIG. 6 is an explanatory diagram showing contents of temporary storage at the time of a recording process.

FIG. 7 is a flowchart of a comparison / report control process relating to the melody.

FIG. 8 is a flowchart showing a comparison / notification control process relating to performance instruction symbols.

FIG. 9 is a flowchart illustrating a sound volume comparison processing routine.

FIG. 10 is a flowchart illustrating a sound length comparison processing routine.

FIG. 11 (a) shows the symbol “pp (Pianissimo)”
(B) is an explanatory diagram of a ranking and correction instruction table in the case of “ff (fortessimo)”.

12A is a diagram illustrating a ranking and correction instruction table for “Staccato”, and FIG. 12B is an explanatory diagram of a ranking and correction instruction table for “Tenuto”.

[Explanation of symbols]

M1 performance information storage means, M2 performance information storage means, M3 performance reproduction means, M4 sample performance information storage means, M5 mismatch mode determination means, M6 correction instruction storage means, M7 correction instruction notification means , M8: notification control means, 1: automatic piano, 3: key,
5 Stepped shutter, 7, 8 Key sensor, 10
... Control unit, 13 ... RAM, 21 ... Floppy disk, 23 ... Floppy disk driver, 25
... key drive solenoid, 27 ... operation panel, 29 ...
Display, 31 ... Sound generator

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Tatsuya Inaba 200 Terashimacho, Hamamatsu-shi, Shizuoka Prefecture Kawai Musical Instruments Manufacturing Co., Ltd. 56) References JP-A-4-199083 (JP, A) JP-A-59-102288 (JP, A) JP-A-3-89492 (JP, U)

Claims (1)

(57) [Claims] 1. Performance information storage means for storing key-depression data and key-release data as performance information along a performance flow, key-depression and key-release timings detected based on key operation positions, Keying data and key release data are created from the keying intensity information calculated based on the speed, and the performance information storage means is stored in the performance information storage means in accordance with the flow of the performance based on the key pressing and key release timing. Means for sequentially outputting individual key press data and key release data in the performance information stored in the performance information storage means in accordance with the flow of the performance, and keying strength information in the output key press data. And a performance reproducing means for reproducing the performance by driving the solenoid with driving power determined based on the key and stopping the driving of the solenoid in accordance with the key release data. Sample performance information storage means for storing data and key release data as model performance information; performance information to be compared and model performance information stored in the performance information storage means for a predetermined checkpoint; Means for comparing the two pieces of performance information to determine whether or not the two pieces of performance information are inconsistent, and correction instruction storage means for preliminarily storing a performance correction instruction for eliminating the inconsistent aspect in accordance with the inconsistent aspect. If there is a discrepancy between the two pieces of performance information, at the check point where the discrepancy exists, a correction instruction notifying means for notifying a corresponding performance correction instruction, and the performance reproducing means reproduces the performance of the comparison target. Sa
Corresponding to the performance timing of the mismatched portion,
A corresponding performance correction instruction is notified by the correction instruction notification means.
Training device according to the automatic player piano, characterized in that it and a notification control means for.