JPH0934452A - Editional device for performance information - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an editional device for performance information which can normalize performance information, while considering performance feature and performance expression of a performer. SOLUTION: In SH4, [(TD-d)/Q+0.5]×Q is computed, and the computed result is received in a register (t). In the formula shown in this SH4, at computing the interior of [], 0.5 is added to it and a decimal fraction is cut off. Consequently, at computing the interior of [], after counting fraction of 0.5 and over as a round unit and cutting away the rest, a decimal point and thereunder are cut off so as to become a positive integer. As shown in the next SH5, 'd' is difference between time information of TD read out from a memory 6 for performance and received and the (t) in which the computed result in the SH4 is received, namely, an error between the stored actual time information and the time information after being quantized. Consequently, when the process from SH2 is repeated, a time value quantized always including the preceding error (d) by the process of SH4 is received in the (t).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a performance information editing device having a so-called quantize function.

[0002]

2. Description of the Related Art Conventionally, a performance information editing device having a quantize function is provided with a memory for storing performance information. When a recording mode is set and a performance is performed on a keyboard, the time information at the time of key depression is recorded. The pitch information, the time information when the key is released, and the pitch information are sequentially stored in the memory. Then, when the quantize switch is operated after the performance information including the time information and the pitch information is stored in the memory in the recording mode as described above, based on each time information, a predetermined unit (quantize unit) is changed from an integral multiple. The normalized time information is calculated, and the time information stored in the memory is rewritten with the normalized time information. This allows
The time information indicating the sounding and silencing timings stored in the recording mode in advance is normalized to an integer multiple of the unit of quantize. For example, the performance information has a plurality of time informations that generate musical tones at the timing of quarter notes. Even if there is an error in the timing of each time information, this error is corrected. Therefore, when the automatic performance is performed using the performance information after the quantization, the note lengths corresponding to the same note are uniformized, and the automatic performance can be performed with accurate performance.

[0003]

As described above, in the conventional performance information editing apparatus having the quantize function,
By rewriting the time information stored in advance at the time of playing to an integral multiple of the unit of quantize, the time information is rewritten. Therefore, the performance characteristics and performance expression of the performer,
For example, for a specific phrase, a performance shorter or longer than the original note length of the note is ignored, and the same note is uniformly edited such that the time intervals from sounding to mute are the same.

The present invention has been made in view of such conventional problems, and provides a performance information editing apparatus capable of normalizing performance information in consideration of performance characteristics and performance expression of a performer. The purpose is to do.

[0005]

In order to solve the above-mentioned problems, according to the present invention, a storage means for sequentially storing time information indicating the timing of sounding and muting and event information for each timing, and this storage. Based on each time information stored in the means, each calculates a normalized time information consisting of an integral multiple of a predetermined unit, in the calculation, means for using the error between the normalized time information and the time information, Rewriting means for rewriting the time information stored in the storage means to the corresponding normalized time information calculated by the calculation means.

In such a configuration, the calculating means calculates the normalized time information which is an integral multiple of the predetermined unit (quantization unit) based on the time information stored in the storage means in advance. In this calculation, the calculation means uses the error between the time information stored in advance in the storage means and the normalized time information. Therefore, the normalized time information calculated by the calculation means is a value that takes into consideration the error between the time information stored in advance in the storage means in the actual performance and the normalized time information.
And all the time information stored in the storage means
The rewriting means rewrites the normalized time information in consideration of this error to complete the editing (quantization).

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. In this embodiment, the present invention is applied to an electronic musical instrument. This electronic musical instrument has
As shown in the block configuration of 1, the SW group 1 and the keyboard 1 are provided. SW group 1 has START / STOP
SW, REC SW, quantize SW, and other function switches are provided. Further, the keyboard 2 is provided with a plurality of keys arranged in the order of pitch and a key switch which is arranged for each key and which is turned on when a key is pressed. The operation information of the keyboard 2 and the SW group 1 is provided. Are taken into the CPU 3.

The CPU 3 has these input information and the ROM 4
It operates based on the programs etc. stored in
While using M5 as a work area, all necessary processing is executed in this electronic musical instrument. That is, CPU
3 executes a process according to a flow described later to perform a data writing process, a reading process, etc. to the performance memory 6, and controls the tone generator 7. The tone generator 7 supplies a tone pitch supplied from the CPU 3. The musical tone waveform is generated based on the data, and the musical tone circuit 8 is configured to generate a musical tone by applying the musical tone waveform to the sound generating circuit 8 including an amplifier and a speaker.

As shown in FIG. 2B, the performance memory 6 is provided with storage areas in the order of serial addresses AD, and time information and event information are alternately stored in each storage area. It The time information is data indicating the timing for reading the event information stored at the next address, and the event information is either an ON flag indicating sounding or an OFF flag indicating mute, as shown in FIG. It is composed of a combination of a heel and pitch data.

Next, the operation of this embodiment having the above configuration will be described with reference to the flowcharts shown in FIG. When the power is turned on, the CPU 3 starts its operation according to the flow shown in FIG. 3, and first, the initialization process (S
Execute A1). In this initialization processing, various registers shown in other flows are cleared or a predetermined initial value is set. Subsequently, switch processing (SA2), keyboard processing (SA3), quantization processing (SA4), reproduction processing (SA5), sound generation processing (AS) which will be described later.
6), and other processing other than these processing (SA7)
And SA2-SA7 while the power is on.
Repeat the loop. Note that, for this main flow, the timer interrupt shown in FIG. 4 is executed interrupting at a constant time interval, and the timer T is counted up (S
B1), therefore T increases at regular time intervals to indicate the current time.

The switch processing (SA2) is performed according to the flow shown in FIG. 5, and START / STOP SW
Process (SC1), REC SW process (SC2), Quantize SW process (SC3), and other SW process (S
C4) is sequentially executed. The START / STOP SW process (SC1) is performed according to the flow shown in FIG. 6. First, it is determined whether or not the START / STOP SW is operated (SD1), and if there is no operation, the process returns to the flow of FIG. If there is an operation, the start flag ST
F is inverted (SD2), the timer T is reset (SD3), and the address register AD is reset (SD4). That is, the timer T counting up by the timer interrupt of FIG.
It is reset every time T / STOP SW is operated, and STF
Is "1" for each operation of START / STOP SW
Repeat "0".

Next, it is judged whether or not this STF is set (SD5). This STF is a flag that indicates a stop when "0" and a start when "1". As a result of the determination in SD5, if STF = 0 and stop is determined, it is further determined whether or not the REC flag RECF is set (SD6). This RE
CF is a flag that is set when the REC SW is operated to start recording in the REC SW process of FIG. 7, which will be described later, and indicates a recording mode by "1" and a reproduction mode by "0".

When the recording mode is stopped when RECE = 1, after the RECF is reset (SD7), the AD value at this time is used as an address and "END" is stored in the performance memory 6. Is stored (SD
8). Therefore, when the recording mode is stopped, "END" is written to the address AD + N at that time, as shown in FIG. Subsequently, by stopping the recording, the timer T does not need to be counted up, so the timer interrupt is prohibited (SD9). If the result of the determination in SD6 is that RECF = 0 and the playback mode is stopped, the flag of the buffer B2, which will be described later, is set to the OFF flag (SD10), and then the timer interrupt is prohibited (SD9). .

On the other hand, as a result of the determination in SD5, STF = 1
If it is a start, the data is read using AD as an address and stored in TM (SD1
1). At this time, AD = 0 by the processing of SD4 described above.
Therefore, at the start time, the time information of the leading area corresponding to AD = 0 is read from the performance memory 6 shown in FIG. 2A and stored in the TM. Subsequently, the prohibition of the timer interrupt is released (SD12), and the timer T is activated.

The REC SW process (SC2) is shown in FIG.
The process is performed according to the flow shown in (1). First, it is determined whether or not the REC SW is operated (SE1), and when there is no operation, the process returns to the flow of FIG. Also, if there is an operation, R
The ECF is set (SE2), and the timer T and the address register AD are reset (SE3). Subsequently, the prohibition of the timer interrupt prohibited by the above-mentioned processing of SD9 is released (SE4). Therefore, as described above, the RECF indicates that the recording mode has been set by operating the RECSW by "1", and
The timer T starts counting up with the start of the recording mode and increases its value. The quantize SW process (SC3) is performed according to the flow shown in FIG. 8. First, it is determined whether or not the quantize SW is operated (SF1), and if there is no operation, the process returns to the flow of FIG. If there is an operation, the quantize flag Q
While setting F, STF and RECF are reset (SF2). Subsequently, the address register AD and the data register d are reset (SF3). Therefore, AD is "0" at the time when the quantize SW is operated by the process of SF3.

On the other hand, the keyboard processing (SA3) in the main flow of FIG. 3 is performed according to the flow shown in FIG. 9, and first, it is detected whether or not the keyboard 2 has been operated (SG1). Then, when the keyboard 2 is not operated and there is no change, the process returns to the main flow without performing the following processing. However, if the keyboard 2 is operated and the key is pressed (ON), the pitch data corresponding to the ON flag and the operation key is stored in the buffer B1 (SG2). In addition, keyboard 2
If there is a key operation and the key is released (OFF), the pitch data corresponding to the OFF flag and the operation key is
It is stored in the buffer B1 (SG3). Therefore, in the buffer B1, whenever the state of the keyboard 2 changes,
A flag indicating the latest change content and pitch data are stored.

Next, it is judged whether or not RECF is set (SG4), and when RECF ≠ 1 and the recording mode is not set, the main flow is immediately executed without performing the subsequent processing. Return to. Therefore, when the keyboard 2 is operated in the state where the recording mode is not set, the writing of data to the performance memory 6 is not performed. However, when RECF = 1 and the recording mode is set, the contents of the timer T are stored in the performance memory 6 by using AD as an address (SG5). At this time, in SE3 of FIG. 7 (REC SW processing) described above, AD is set to “0” when the REC SW is operated, and the timer T is reset in SE3 and then SE4.
Has started counting up. Therefore, REC
When pressing the keyboard 2 for the first time after operating the SW, this S
By the processing of G5, the value of the timer T when the key is first pressed is used as the time information, and the address "0" of the performance memory 6
Will be written.

Subsequently, the ON flag or the OFF flag and the pitch data are stored in the performance memory 6 using (AD + 1) as an address (SG6). By the processing of SG6, the performance memory 6 is placed in the area next to the time information stored in SG5, together with the ON flag if the key is pressed and the OFF flag if the key is released, and the pitch data corresponding to the key. Will be written. And next SG7
Now, by the processing of SG5 and SG6, the performance memory 6
Since the data has been written in the areas corresponding to AD and AD + 1 at, the AD value is incremented by adding "2" to the current AD value, and then the process returns to the main flow of FIG. Therefore, in this main flow,
By repeatedly executing the loop of SA2 to SA7, in the recording mode, as shown in FIG. 2A, time information and event information including a combination of an ON flag or an OFF flag and pitch data are recorded. Are written alternately. Further, when the START / STOP SW is operated to stop, the S in FIG.
At D8, "END" is written in the address of the performance memory 6 indicated by AD at that time, and the recording is completed.

Further, the quantization process (SA4) in the main flow of FIG. 3 is performed according to the flow shown in FIG. 10, and it is judged whether or not the quantize flag QF is set (SH1). Returns to the main flow without executing the subsequent processing. Here, QF is a flag that is set by operating the quantize SW in SF2 in FIG. 8 described above. Therefore, the processing after SH2 shown in FIG. 10 is started on condition that the quantize SW is operated.

When the quantize SW is operated and QF = 1, AD is set as an address and
Data is read from the performance memory 6 and stored in TD (AH2). At this time, AD is performed in SF3 of FIG. 8 described above.
Is reset, the time information stored in the head area of the performance memory 6 is read and stored in the TD at the start of the quantize.
Next, it is determined whether or not the data stored in this TD is time information (SH3), and if it is not time information, SH3
To SH7. Then, in this SH7, it is judged whether or not the stored data is "END", and "E"
If the event information is not "ND", AD is incremented (SH9), and then the processing from SH2 is repeated. If the data stored in the TD is time information as a result of the determination in SH3, [(TD-d) /
Q + 0.5] × Q is calculated, and the calculation result is stored in the register t (SH4). In the expression shown in SH4, 0.5 is added in the calculation in [], and the decimal part is truncated. Therefore, in the operation in [], the first decimal place is rounded off, and then the fractional part is rounded down to a positive integer. Further, d is the difference between the time information of the TD read and stored from the performance memory 6 and the time t at which the calculation result in this SH4 is stored, that is, the stored actual time, as shown in SH5. Is the error between the time information of and the normalized time information after quantization. Therefore, in the calculation in SH4, since "TD-d" exists, the time value quantized in consideration of the previous error d is always stored in t when the process from SH2 is repeated. It will be.

That is, in the processing of SH4, assuming that the calculation is performed by [TD / Q + 0.5] × Q without using this error d, the timing of reading the next event information is the same as in the conventional apparatus. It will always depend on TD / Q. Therefore, if each event information is stored at each time of a time interval that is close to a quarter note length, the time value of each same time interval is stored in t by the processing of this assumed SH4. However, in the present embodiment, in the calculation of SH4, "TD-
As described above, the time value quantized by taking the error d into consideration is stored in t as described above. In addition, when the error d is present, the performance is performed in the recording mode for performance. Since it is caused by the performance characteristics of the person who has stored the performance result in the memory 6 and the peculiar performance expression,
The characteristics of the performer can be added to the time information that is an integer multiple of the quantised Q.

As described above, if the time information consisting of an integer multiple of Q normalized by the quantize taking into account the error d is stored in t, AD is used as an address and t
The content of is stored as time information (SH6). Subsequently, after incrementing AD (SH9), the processing from SH9 is repeated. Therefore, each time information stored in the performance memory 6 in the recording mode is SH3.
By the processing from SH6 to SH6, the time is rewritten into the quantized time information in consideration of the error d. Then, after the last time information stored in the performance memory 6 is quantized, END is read and stored in TD,
After proceeding to SD3 → SD7 → SD8 and resetting the QF, the process returns to the main flow.

The reproduction process (SA5) is performed according to the flow shown in FIG. 11, and it is determined whether STF is set (SI1). If STF = 0 and the stop state is set, the reproduction process is immediately executed. Return to the main flow. Also, ST
When F = 1 and the start state is reached, as shown in FIG.
Timer T reset at the start of SD3
Determines whether or not the value of has reached the time indicated by the time information stored in the register TM in SD11 of FIG.
2). Then, T = TM, and the value of the timer T becomes TD.
When the time indicated by the time information is reached, AD is incremented (SI3), and the incremented AD is used as an address to read data from the performance memory 6 (SI4).

Next, it is identified whether the read data is event information, time information, or "END" (SI5). As a result of this identification, the read information is ON
If the event information is a combination of the flag or the OFF flag with the pitch data, the read data is stored in the buffer B2 (SI6) and then SI3.
Repeat the process from. If the data read out in SI4 is time information when the process from SI3 is executed again, the read out data is stored in the register TM (SI7), and then the process returns to the main flow. At this time, as shown in FIG. 2A, since the time information and the event information are alternately stored in the performance memory 6, the loop of SA2 to SA7 of the main flow is executed once. Event information is always stored in the buffer B2 (SI6) and the register TM
The time information is stored in (SI7).

When the loop of SI2 to SI7 is repeatedly executed, each time the value of the timer T reaches the time information stored in TM by the previous processing and T = TM,
Processing after SI3 is executed. Therefore, the event information stored in the performance memory 6 is sequentially read out at the timing of the time information stored in the performance memory 6 and stored in the buffer B2. At this time, if the quantize SW has already been operated after recording and the quantization processing of FIG. 10 described above has been completed, the event information will be sequentially read out at the timing indicated by the quantized time information. Also, after recording, quantize S
When W is not operated and the quantization processing of FIG. 10 is not completed, the event information is sequentially read at the timing indicated by the time information stored in the performance memory 6 during recording. Then, after all the event information and the time information have been read, when "END" of the final data is read, the process proceeds from SI5 to SI8 to reset the STF, and then disables the unnecessary timer interrupt ( S
I9), and returns to the main flow.

In the sounding process (SA6), as shown in FIG. 12, a keyboard sounding process (SJ1),
The automatic performance process (SJ2) is executed. The keyboard sounding process (SJ1) is performed according to the flow shown in FIG. 13, and the pitch data of the buffer B1 is supplied to the sound source 7 (SK
1). That is, as described above in the keyboard processing of FIG. 9, when the keyboard 2 is operated and the key is pressed (ON), the pitch data corresponding to the ON flag and the operation key is stored in the buffer B1. If it is stored (SG2) and the key is released (OFF), the pitch data corresponding to the OFF flag and the operation key is stored in the buffer B1 (SG).
2).

Therefore, in SK1, the pitch data stored in the buffer B1 is supplied to the sound source 7, and in the next SK2, the flag stored in the buffer B1 is either an ON flag or an OFF flag. Determine whether. If the result of this determination is that the key is an ON flag indicating key depression, a sounding command is sent to the sound source 7 (SK3). As a result, the sound source 7 starts to generate a musical tone waveform based on the supplied pitch data. When the generated musical tone waveform is applied to the tone generating circuit 8 including an amplifier and a speaker, the tone generating circuit 8 starts to generate a tone having a pitch corresponding to the operation key.

If the result of determination in SK2 is an OFF flag indicating key release, a mute command is sent to the sound source 7 (SK4). As a result, the sound source 7 rapidly attenuates the musical tone waveform corresponding to the supplied pitch data. Along with the rapid attenuation of the musical tone waveform, the musical tone of the pitch corresponding to the operation key generated from the tone generation circuit 8 is silenced. That is, the above-described keyboard processing of FIG. 9 and the keyboard sounding processing shown in FIG. 13 are repeatedly executed in the loop of SA2 to SA7 of the main flow, so that the sound corresponding to the operation key is generated in response to the key depression. A high-pitched tone is generated, and the tone of the pitch corresponding to the operation key is muted with the release of the key.
You can play by operating.

On the other hand, the automatic performance sound generation processing (SJ2) is performed according to the flow shown in FIG. 14, and the pitch data in the buffer B2 is supplied to the sound source 7 (SL1). That is, FIG.
As described above in the reproduction process of No. 1, SA2 of the main flow
Every time the loop of SA7 is executed once, the event information is always stored in the buffer B2 (SL6). Therefore, in SL1, the pitch data stored in this buffer B2 is supplied to the sound source 7, and the next SL2
Then, it is determined whether the flag stored in the buffer B2 is an ON flag or an OFF flag. If the result of this determination is that the key is an ON flag indicating key depression, a tone generation command is sent to the sound source 7 (SL3). As a result, the sound source 7 starts to generate musical tone waveforms based on the supplied pitch data, and the tone generation circuit 8 starts to generate musical tones corresponding to the pitch data. If the result of determination in SL2 is an OFF flag indicating key release, a mute command is sent to the sound source 7 (SL4). As a result, the sound source 7 rapidly attenuates the musical tone waveform corresponding to the supplied pitch data. Along with the rapid attenuation of the musical tone waveform, the musical tone with the pitch corresponding to the pitch data generated from the tone generation circuit 8 is muted. Therefore, the sound generation command (SL3) and the mute command (SL4) of this automatic performance sound generation process are executed each time the loop of SA2 to SA7 of the main flow is repeated, whereby the automatic performance advances.

At this time, as described above, if the quantize SW is operated and the quantization processing of FIG. 10 is completed, at the timing indicated by the quantized time information,
Since the event information is sequentially stored in the buffer B2, sound generation and mute are executed at the timing indicated by the time information after the quantization. In the time information after the quantizing, the characteristics of the person who stored the performance in the performance memory 6 and the peculiar performance expression are added by the processing of SH4 in FIG. Even if the data is reproduced, it is possible to perform automatic performance in which the performance expression is not uniform. If the recording is not quantized, the event information is sequentially read at the timing indicated by the time information stored in the performance memory 6 during recording, and the performance during recording is reproduced by automatic performance.

FIG. 15 shows the processing contents of the quantization processing (SA4) according to another embodiment of the present invention.
First, it is determined whether the quantize flag QF is set (SM1). Here, QF is a flag that is set by operating the quantize SW in SF2 in FIG. 8 described above. Therefore, this FIG.
The processing after SM2 shown in (1) is started on condition that the quantize SW is operated. Then, when QF = 1 due to the operation of the quantize SW, each register TD, AD, AD1, N used in this flow is reset (SM2), and then the value of the address register AD is set to AD1 (SM3. ). Next, using the value indicated by this AD1 as an address, data is read from the performance memory 6 (SM4), and it is determined whether the read data is time information (SM5). As a result of this determination,
If the read data is time information, the read data is added to the current value of TD to update the value of TD (SM6).

Next, after incrementing N (SM
7), it is determined whether N = n (SM8). Here, “n” is the other S in SC5 of FIG. 5 by operating the numerical keys provided in the switch group 1.
A value set in the W process or a predetermined value stored in the ROM 4 in advance, for example, "2". Therefore, if the present time is immediately after the start of this quantization processing, N = 1 and N ≠ n by the processing of SM7, and therefore, the processing advances from SM8 to SM9 and AD1 is decremented. Next, it is determined whether the value of this decremented AD1 is less than 0 (SM10). At this time, immediately after the start of this quantization processing, AD1 = 0 by the processing of SM3 and AD1 by the processing of SM9.
= -1. Therefore, in this case, the process proceeds from SM10 to SM11, divides the value of TD updated in SM6 by the value of N, and stores the value in TD (SM11). Subsequently, the same processing as in SH4 to SH6 described above in FIG. 10 is performed in SM12 to SM14. This allows SM
The time value quantized by adding the error d by the processing of 12 is stored in t, and the time information of the performance memory 6 is rewritten by the processing of SM14 by the quantized time information by adding the error d. . Continued
After incrementing AD (SM15), N is reset (SM16), and the processing from SM3 is repeated.

Then, when the data read in SM4 is not the time information when the processing from SM3 is repeated, the process proceeds from SM5 to SM17 and the data is "EN".
If it is the event information instead of "END", the process proceeds from SM17 to SM9, and AD1 is decremented. At this time, the increment in SM15 has already been executed multiple times, and for example, AD
= 10, AD1 is processed by SM3.
= 10. Therefore, even if AD1 is decremented by SM9, it will not be AD1 <0.
The determination of SM10 is NO, and SM10 to SM4
Then, the data is read using the decremented AD1 as an address.

If the read data is not time information, the process proceeds to SM5 → SM17 → SM9 to decrement AD1 and then S10 via SM10.
Return to M4. Then, if the time information is read in SM4, after performing the above-described processing of SM6 and SM7,
It is determined whether N = n (SM8). As a result of this determination, when N = n, that is, n = 2 as in this example.
If SM6 is processed twice, then S
The process proceeds from M8 to SM11, and the processes after SM11 are performed. Therefore, SM12 to SM1 are based on the time information that is the average value of the time information indicated by this AD and the time information before this (in this example, the time information up to two times before).
Thus, the processing of No. 4 is performed, which allows quantizing while preserving the performance characteristics in the recording mode.

[0035]

As described above, according to the present invention, when the normalized time information which is an integral multiple of the predetermined unit is calculated based on each time information included in the performance information, the normalized time information and the time information are combined. The error is used. Therefore, the normalized editing can be performed while reflecting and adding the performance characteristics and performance expression of the performer without ignoring them. Further, by sequentially using the error between the time information and the corresponding normalized time information when calculating the normalized time information corresponding to the next time information, the performance characteristics of the performer over the entire performance information can be determined. Can be reflected. further,
With the configuration having the sounding instructing means for sequentially reading out the normalized time information and the event information and instructing the sounding, it becomes possible to reproduce the performance in which the performance characteristics and the like are reflected while ensuring the accuracy of the note length.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of an electronic musical instrument to which an embodiment of the present invention is applied.

FIG. 2 is a diagram showing stored contents of a performance memory.

FIG. 3 is a flowchart showing a main flow.

FIG. 4 is a flowchart showing a timer interrupt.

FIG. 5 is a flowchart showing the contents of switch processing.

FIG. 6 is a flowchart showing the contents of START / STOP processing.

FIG. 7 is a flowchart showing the contents of REC SW processing.

FIG. 8 is a flowchart showing the contents of a quantize SW process.

FIG. 9 is a flowchart showing the contents of keyboard processing.

FIG. 10 is a flowchart showing the contents of quantization processing.

FIG. 11 is a flowchart showing the content of a reproduction process.

FIG. 12 is a flowchart showing the contents of a sound generation process.

FIG. 13 is a flowchart showing the contents of keyboard sounding processing.

FIG. 14 is a flowchart showing the contents of automatic performance sounding processing.

FIG. 15 is a flowchart showing the contents of quantization processing in another embodiment of the present invention.

[Explanation of symbols]

 1 SW group 2 keyboard 3 CPU 4 ROM 6 performance memory 7 sound source

Claims (3)

[Claims] 1. Storage means for sequentially storing time information indicating timings of sounding and muting and event information for each timing, and an integral multiple of a predetermined unit based on each time information stored in the storage means. Computation means that respectively computes the normalized time information and uses an error between the normalized time information and the time information at the time of computation, and a correspondence obtained by computing the time information stored in the storage means by the computing means. And a rewriting means for rewriting the normalized time information to the performance information editing apparatus. 2. The calculating means uses an error between the time information and the corresponding normalized time information when calculating the normalized time information corresponding to the next time information. The performance information editing device described. 3. A sounding instructing means for sequentially reading out the normalized time information and the event information from the storage means rewritten by the rewriting means and instructing sounding is further provided. Alternatively, the performance information editing device described in 2.