JP3316530B2 - Music control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a musical tone control device used for electronic musical instruments and the like.

[0002]

2. Description of the Related Art Conventionally, there are electronic musical instruments using various sound sources such as a PCM system and an FM system.

In musical tone waveform data generated in such an electronic musical instrument, tone generation characteristics are set according to the type of musical tone to be emitted. For example, musical sound waveform data having a tone color such as a wind instrument or a guitar is generated so as to include a noise (noise) component in a sound generation start portion (attack portion). It is generated so that the attack part does not include noise (noise) components.

Further, musical tone waveform data having a tone color of a piano or the like is generated such that the rise of the waveform of an attack portion is relatively fast. Conversely, musical tone waveform data having a brass tone color is generated in the attack portion. Is generated such that the rising of the waveform of the waveform is relatively slow.

[0005]

In the case where musical tones are generated based on musical tone waveform data having various sound generating characteristics as described above, if the playing state is different, a preferable sound is not produced depending on the type of musical sound to be generated. May not be.

[0006] For example, a case where a player performs at a high performance speed. In such a case, for example, when musical tones based on musical tone waveform data of a tone containing much noise in an attack portion are successively generated in response to a fast performance, mute (note off) from the start of each musical tone (note on). )
Since the time until the time becomes short, only the noise component of the attack part is output from the musical tone waveform data corresponding to each musical tone. As a result of generating musical tones based on such musical tone waveform data, there is a problem in that the performance becomes inconspicuous as a whole with no noise, and the musical tones having the original timbre cannot be generated. ing.

Further, for example, when the tones based on the musical tone waveform data of the tone having a slow rise of the waveform of the attack portion are successively generated in response to the fast performance, the time from note-on to note-off of each musical tone is also short. Therefore, as for the musical tone waveform data corresponding to each musical tone, only the waveform portion having almost no amplitude level in the attack portion is output. As a result of the musical tone being generated based on such musical tone waveform data, a performance in which a tone similar to a musical tone can hardly be heard as a whole, and in this case also, a musical tone having an original tone cannot be generated. Have a point.

Thus, in a conventional electronic musical instrument or the like,
If the performance state is different, there is a problem that a preferable tone is not produced depending on the type of musical tone to be emitted. SUMMARY OF THE INVENTION An object of the present invention is to make it possible to generate a stable and stable musical tone regardless of the performance state.

[0009]

According to the present invention, a musical tone generated within a predetermined time in the past is changed from note-on to note-off.
Performance state evaluation means for evaluating the performance state based on the average value of the time up to. This means, for example, within a predetermined time in the past from the timing when a new note-on instruction is issued
The performance speed is evaluated from the average value of the gate time of the musical tone that is pronounced at the time .

Next, there is provided a tone characteristic control means for controlling the characteristic of a tone generated based on a new note-on instruction based on the performance state evaluation result by the performance state evaluation means. This means controls, for example, a characteristic of a rising portion of a tone generated based on a new note-on instruction, based on a performance speed evaluation result by the performance state evaluation means. More specifically, the musical tone control means shifts a waveform generation start position of a musical tone generated based on a new note-on instruction based on a performance speed evaluation result by the performance state evaluating means. Alternatively, the tone control means changes the amplitude envelope characteristic of the rising portion of the tone generated based on the new note-on instruction based on the performance speed evaluation result by the performance state evaluation means.

[0011]

According to the present invention, the performance state evaluating means is used.
From the note-on of the tone that occurred within the predetermined time in the past
Performance status is evaluated based on the average value of the time until note-off.
Based on the result of the evaluation.
Of the tone generated in response to a new note-on instruction
Control gender .

[0012]

As a result, a preferred musical tone corresponding to the currently recognized performance state is generated. Specifically, for example,
The performance state evaluation means evaluates the performance speed as an average value of the time from note-on to note-off of a tone generated during a predetermined period in the past, and the tone characteristic control means evaluates a new note based on the evaluation result. The characteristics of the rising portion of the tone generated based on the ON instruction, for example, the waveform generation start position of the tone and the amplitude envelope characteristic of the rising portion of the tone are changed.

As described above, for example, in the case of a tone having a noise component in the rising portion, when the performance speed increases, the waveform generation start position of the musical tone is shifted so that the noise component in the rising portion is not generated, and the noise component is reduced. The removed musical tone is pronounced.

[0015] For example, in the case of a tone having a slow rising waveform, the amplitude envelope characteristic of the rising portion of the musical tone is changed so that the rising waveform has a fast rise and its maximum level also increases.

As described above, the waveform generation start position or the amplitude envelope characteristic of the rising portion is changed, so that a preferred musical tone corresponding to the currently recognized playing speed is generated.

[0017]

Embodiments of the present invention will be described below with reference to the drawings. <Structure of Embodiment> First, the structure of an embodiment of the present invention will be described.

FIG. 1 is an overall configuration diagram of an embodiment of the present invention, which is realized as an electronic musical instrument. First, the electronic musical instrument 10
Is a CPU (Central Processing Unit) that controls the entire device
1, a switch unit 2 connected to the CPU 1, a low-pass filter LPF4, an amplifier (amplifier) 5 connected to the low-pass filter LPF4, and a speaker 6 connected to the amplifier 5.

The switch section 2 is a musical instrument operation input section including a keyboard switch, a function switch, and the like (not shown), and outputs performance information input by a user's switch operation to the CPU 1.

The CPU 1 has a ROM (Read Only Mem
ory) 1a, a RAM (Random Access Memory) 1b, a latch 1c, a latch 1d, and a D / A (digital / analog) converter 1e.

The ROM 1a is a control data / waveform ROM.
In addition to various tone control programs for realizing the operation of the operation flowchart described later, various tone waveform data and tone control data (described later in FIG. 2) such as an envelope plate and an envelope level are stored. ing.

The RAM 1b includes, for example, a tone generation channel area for storing tone control data corresponding to each of the sixteen tone generation channels, and a tone waveform data accumulation register R.
Etc.

The CPU 1 executes software sound source processing for 16 sounding channels based on performance information input from the switch unit 2 and according to a program in the ROM 1a, using the tone control data in the ROM 1a and the RAM 1b. As a result, the tone waveform data accumulated in the tone waveform data accumulation register R of the RAM 1b is output to the latch 1c.

The latch 1c is generated by the above-described sound source processing in response to a sound source processing end signal COM which is input each time the sound source processing for 16 sounding channels ends from a control unit including an instruction decoder and the like (not shown) inside the CPU 1. The tone waveform data for one sample is latched.

Here, the time required for the sound source processing is as follows.
Since the sound source processing changes depending on the execution condition of the software for sound source processing, the timing at which the sound source processing ends and the musical tone waveform data is latched in the latch 1c is not constant. Therefore, the latched tone waveform data cannot be directly output to the D / A converter 1e.

Then, the latch 1d latches the tone waveform data output from the latch 1c by an interrupt signal INT synchronized with a sampling clock having a frequency of 44 kHz (kilohertz) generated by dividing the frequency by a hard clock, The output is supplied to the D / A converter 1e.

As a result, the D / A converter 1e converts the above-mentioned tone waveform data into an analog tone signal at a timing accurately synchronized with the sampling clock, and outputs the analog tone signal to the low-pass filter 4.

The low-pass filter 4 smoothes the access tone signal described above, and the amplifier 5 amplifies the signal and emits the sound to the outside via the speaker 6. <Structure of Tone Control Data> FIG. 2 shows an example of tone control data stored in the ROM 1a.

The tone control data is stored in pairs for each tone. FIG. 2 shows the tone control data for one tone.
Only those relevant to the present invention are shown.

The data area 2a stores amplitude envelope attack rate data, amplitude envelope attack level data, start address data, and the like. The amplitude envelope attack rate data is data that determines the rising speed of the attack portion in the amplitude envelope of a musical tone. Amplitude envelope attack level data is the amplitude envelope of the musical tone.
This is data for determining the maximum amplitude level of the attack section.
The start address data is the start address of the musical tone waveform data to be read from the waveform data area in the ROM 1a, and the readout waveform is determined by this data, and the timbre is determined.

In the data area 2b, amplitude envelope attack rate shift data, amplitude envelope attack level shift data,
Shift data such as start address shift data is stored. These data are data particularly relevant to the present invention.

The amplitude envelope attack rate shift data is data for shifting the rising speed of the attack portion in the amplitude envelope of the musical tone so that the rising speed of the tone envelope is increased when the playing speed is high. Further, the amplitude envelope attack level shift data is data for shifting the maximum amplitude level of the attack portion in the amplitude envelope of the musical tone so as to increase when the playing speed is high. The values of these shift data are determined according to the magnitude of the value of the amplitude envelope attack rate data, that is, depending on whether the rise of the waveform of the attack part is fast or slow, and when the rise of the waveform of the attack part is slow. Is set to a large value.

The start address shift data is provided so that when the playing speed is high, the noise component of the attack portion is not generated in a tone having a noise component in the attack portion.
This is data for shifting the start address of the tone waveform data to be read from the waveform data area in the ROM 1a. For a tone color that does not include a noise component in the attack portion, the value of this data is set to 0 and the address is not shifted. <Data Structure of Register> Next, among the registers provided in the RAM 1b, the data structure of those related to the present invention is shown in FIGS. 3 (a) and 3 (b) and FIGS. 4 (a) to 4 (f).

The 16 gate-on times & shown in FIG.
Note-off time register GOF (i) (i = 1 to 16)
Is a 32-bit memory that stores a 16-bit gate-on time i and a 16-bit note-off time i, which are the time from the note-on time to the note-off time, corresponding to each of the 16 sounding channels. Register.

Each of the sixteen note-on time registers NOT (i) shown in FIG. 3B is a 16-bit register for storing a note-on time corresponding to each tone generation channel among the sixteen tone generation channels. .

Therefore, the registers GOF (i) and NOT
According to (i), three data of a note-on time, a note-off time, and a gate-on time are stored for each of the 16 sounding channels.

The time counter TC shown in FIG. 4A is a 16-bit register for measuring the note-on time and the note-off time. The time counter TC is 1 msec (millisecond) by a timer interrupt routine 1 shown in FIG. It is incremented by +1 each time. In this case, when the count value is incremented after reaching 65535 (decimal notation), it returns to 0. Therefore, the time counter TC is:
The value from 0 to 65535 is counted repeatedly.

The channel register CH shown in FIG.
Gate-on time & note-on time register GOF (i)
Alternatively, when the note-on time register NOT (i) is accessed, 4 is used to designate a sounding channel.
Bit register.

The average gate-on time register in FIG. 4C is a 16-bit register that stores the average value of the gate-on time. The latest note-on time register NNT in FIG. 4D is a 16-bit register that stores the latest note-on time.

The gate-on time ignoring flag GOI shown in FIG. 4E is a 16-bit flag register. One bit corresponds to one sounding channel, the MSB is the first channel,
LSB corresponds to the 16th channel.

In this flag register, when note-on occurs continuously, if the interval of the note-on is determined to be extremely short and it is determined that the consecutive note-on constitutes a chord, the first note constituting the chord is generated. Is set to a bit corresponding to a sound channel to which a sound other than the note-on sound is assigned. In the embodiment of the present invention, as will be described later, it is necessary to determine a state in which the performance is performed at a high speed and the interval between note-on becomes short. In order to distinguish this state from the state in which the chord is played, The gate on time ignoring flag GOI is used. That is, the interval of note-on when a chord is played becomes even shorter than the interval of note-on when a performance is performed at a high speed. Therefore, by discriminating such a state, the gate-on time ignoring flag GOI is determined. Is set.

Finally, the working register W shown in FIG. 4 (f) temporarily stores intermediate data and the like during the operation.
It is a 6-bit register. <Schematic Operation of Embodiment> Next, the schematic operation of the embodiment of the present invention will be described.

First, when one note-on is assigned to one sound channel i, the note-on time is stored in the note-on time register NOT (i) corresponding to the sound channel i.

Next, when note-off occurs for one sounding channel i, the note-off time is stored in the gate-on time & note-off time register GOF (i) corresponding to the sounding channel i, and the sounding channel i Is calculated as a gate-on time from the note-on time stored in the note-on time register NOT (i) corresponding to the note-on time to the above-mentioned note-off time, and this is calculated as the gate-on time & note-off time register GOF corresponding to the sound channel i. (I).

In this manner, the note-on time, note-off time, and gate-on time are sequentially stored for each of the plurality of tone generation channels. When a new note-on occurs, the gate-on time & note-off time register GOF (i) (i = 1 to 16) extracts all gate-on times whose note-off time is within a predetermined time in the past from the current time. , The average value of the extracted gate-on times is calculated. The average value of the gate-on time corresponds to the performance speed from the last past to the present.
It is slow if it is long.

Therefore, the average value of the calculated gate-on time is input to a conversion table having the conversion characteristics as shown in FIG. 9, so that the performance speed data corresponding to the gate-on time is calculated as an output. Is done.

Subsequently, the value of the calculated performance speed data is stored in the data area 2 shown in FIG.
b is multiplied by the amplitude envelope attack rate shift data, the amplitude envelope attack level shift data, and the start address shift data, respectively, and the result of the multiplication is a data area 2a shown in FIG.
Is added to the amplitude envelope attack rate data, the amplitude envelope attack level data, and the start address data stored in
The rising rate of the amplitude envelope in the attack section, the maximum level, and the start address of the musical tone waveform data are varied.

Now, in the case of a tone having a slow rise in the waveform of the attack portion, large values are set as the amplitude envelope attack rate shift data and the amplitude envelope attack level shift data. A large value is added to the envelope attack rate data and the amplitude envelope attack level data. As a result, a tone with a rapid rise of the amplitude envelope in the attack portion and a large maximum level in that portion is generated.

On the other hand, in the case of a tone color containing a noise component in the attack portion, a large value is set as the start address shift data.
A large value is added to the start address data. As a result, the noise component of the attack portion is not generated so that R
The start address of the musical tone waveform data to be read from the waveform data area in the OM1a is shifted, and a musical tone from which a noise component has been removed is generated.

As described above, the waveform readout position or the amplitude envelope characteristic of the attack portion is changed, and as a result,
A preferred tone corresponding to the currently recognized playing speed is produced. <Specific Operation of Embodiment> Next, a specific operation of the embodiment of the present invention will be described.

The main flowchart of FIG.
In particular, this is a process other than the sound source process executed by the CPU 1 in a state where an interrupt is not applied from an interrupt control unit (not shown), and is realized as a process in which the CPU 1 executes a control program stored in the ROM 1a.

In FIG. 5, first, in step S1, the power is turned on, and the initial settings such as the contents of the RAM 1b (see FIG. 1) in the CPU 1 are performed. In this initial setting, all "1" are set to the first 16 bits indicating the gate-on time of the gate-on time & note-off time register GOF (i). As a result, in the initial state, the average value of the gate-on time is increased, so that it is possible to avoid the inconvenience of erroneously determining that the performance speed is fast from the beginning.

The last 16 bits indicating the note-off time are all set to "0". Further, as in the case of the gate-on time described above, "1" is set to all 16 bits in the average gate-on time register AGO so that it is not erroneously determined that the playing speed is fast from the beginning.

Other register groups, that is, a note-on time register NOT (i), a time counter TC, a tone generation channel register CH, a latest note-on time register NNT,
“0” is set in the gate-on time ignoring flag GOI and the working register W.

Next, in step S2, the state of each switch such as a function switch of the switch unit 2 connected to the outside of the CPU 1 is scanned, and the state of each switch is taken into a key buffer area in the RAM 1b.

In step S3, as a result of the above-described scanning, the function switch whose state has changed is identified, and the processing of the corresponding function is performed. For example, if the tone color changeover switch is operated by the user, the access address of the tone control data accessed in the ROM 1a as shown in FIG. 2 is changed.

In step S4, the state of the key switch is taken in the keyboard of the switch unit 2 in the same manner as in the case of the function switch, and in the subsequent step S5, the changed key switch state, for example, the key is released from the released state. When it is determined that the key is released or the key is released from the key-depressed state, the corresponding processing is executed. The processing in step S5 is most relevant to the present invention, and this processing will be further described later.

At step S6, an envelope is added to the pitch of the tone waveform data of the sound channel to be sounded, and pitch data is set in the corresponding sound channel area on the RAM 1b. The process of changing the state of the sounding channel to which the note has been turned off in step S5 while the sound is muted is executed.

Next, S5 of the main flowchart of FIG.
The note-on process in the above process will be described with reference to the flowchart of FIG. In the following description, the symbolic name of each register indicates the value of that register as it is.

In step S601, it is determined whether a new note-on has occurred. If a new note-on does not occur, the determination result of step S601 is NO, and the note-on process is not executed.

If a new note-on occurs, it is determined in step S602 whether or not there is an empty channel among the 16 sounding channels. If there are no free channels,
The decision result in the step S602 is NO, and the note-on process is not executed.

If there is a free channel, step S603
Then, a channel assignment process for selecting one of the available channels is executed. Further, in step S604, the number of the assigned sounding channel is stored in the sounding channel register CH.
Is set to

Next, in step S605, the time value of the latest note-on time register NNT (previous note-on time) is subtracted from the current time value indicated by the time counter TC, and the time difference (TC-NNT) resulting from the subtraction is obtained. Is determined to be negative.

Here, the current time value of the time counter TC is counted up at regular intervals by a program for executing the timer interrupt routine 1 of FIG. Now, an interrupt control unit (not shown) in the CPU 1 interrupts the program corresponding to the main flow in FIG. 5 every 1 msec, interrupts the processing of the program, and corresponds to the timer interrupt routine 1 in FIG. The execution of the processing of the program to be started is started. In the timer interrupt routine 1 of FIG. 7, in step S701, the time counter T in the RAM 1b
The value of C is incremented by one. In addition, the timer interrupt routine 2 of FIG.
Interrupt priority is high.

As described above, every time 1 msec elapses, the current time value of the time counter T increases by one. The time counter TC is a 16-bit register, as described above with reference to FIG. 4, and the count value counted by this counter is incremented after reaching 65535 (decimal notation). Return to 0. Therefore, the time counter TC repeatedly counts values from 0 to 65535. The operation is shown in FIG.
In FIG. 8, 0 to 65535 in decimal notation corresponds to 0 to FFFF in hexadecimal notation.

Now, when the time difference (TC-NNT) is negative, as shown in FIG. 8A, the maximum value "FFFF (H)" of the time counter TC ((H) is added. The value of the time counter TC is added to a value obtained by subtracting the time value of the latest note-on time register NNT from the value “10000 (H)”, which is one greater than the value of “10000 (H)”. The value obtained is the true time difference between the current time and the latest note-on time. That is, step S605
If the determination result in step S606 is YES, step S606 is performed.
The true time difference is calculated and stored in the working register W.

On the other hand, when the time difference (TC-NNT) is positive, the value obtained by subtracting the time value of the latest note-on time register NNT from the current time value of the time counter TC is the current time and the latest note value. This is the true time difference from the ON time. That is, if the decision result in the step S605 is NO, the true time difference is calculated in a step S607, and the calculated time difference is stored in the working register W.

Here, the unit of time is msec (millisecond) as described above. Next, step S608
In, the time value of the latest note-on time register NNT is rewritten to the current time value of the time counter TC (the current note-on time value).

Subsequently, in step S609, it is determined whether or not the time difference between the current time obtained in the working register W and the latest note-on time is less than 32 msec in decimal notation. By this determination processing, it is determined whether or not a chord has been played. That is, in the constituent sounds of one chord, if the constituent sounds other than the constituent note to which the note is first turned on are continuously turned on in a very short time, the determination result in step S609 becomes YES.

If the decision result in the step S609 is YES, in a step S610, "1" is set to a bit corresponding to the current sounding channel set in the sounding channel register of the gate-on time ignoring flag GOI, and the process in the step S611 is performed. Is not executed. As a result, among the constituent tones of one chord, the constituent tones other than the one to which the note is first turned on are not stored in the note-on time even if the sound generation processing is performed by the note-on, and the note-off time in FIG. And the gate-on time is not stored. Thus, it is possible to prevent a harmony performance from being erroneously determined to be a performance having a high performance speed.

On the other hand, the time difference between the current time obtained in the working register W and the latest note-on time is 32 ms.
c or more, and the determination result of step S609 is YES
, The note-on time register NOT corresponding to the current sounding channel i set in the sounding channel register
(I) stores the time value of the latest note-on time register NNT (the current note-on time).

Next, at step S612, the gate on time & note off time register GOF (i) (i = 1 to
In 16), all the gate-on times whose note-off time is within the past 1024 msec in decimal notation from the time value (current note-on time) of the latest note-on time register NNT are extracted, and the average value of the extracted gate-on times Is calculated and stored in the average gate-on time register AGO. The contents of the gate-on time & note-off time register GOF (i) will be described later with reference to FIG.
Is set in the note-off process. The above average value of the gate-on time corresponds to the performance speed from the last past to the present. The shorter the gate-on time, the faster the performance speed, and the longer the gate-on time, the slower the performance speed.

Subsequently, in step S613, the ROM 1a
From the data area corresponding to the currently set tone,
As the tone control data, the amplitude envelope attack rate data, the amplitude envelope attack level data, and the start address data shown in FIG. 2 are read out, and as the shift data, the amplitude envelope attack rate shift data, the amplitude shown in FIG. The envelope attack level shift data and the start address shift data are read. Then, the performance data obtained by converting the time value of the average gate-on time register AGO with the shift table stored in the ROM 1a and having the characteristics shown in FIG. 9 is multiplied by each of the above-mentioned shift data, and the multiplication result is obtained. Is added to the tone control data corresponding to each shift data, and the addition result is stored in the RAM.
This is set in the sounding channel area corresponding to the current sounding channel set in the sounding channel register in 1b.

More specifically, the performance speed data output from the shift table is multiplied by the amplitude envelope attack rate shift data, the result of the multiplication is added to the amplitude envelope attack rate data, and the addition result is corrected for the corrected amplitude. This is set in the sound channel area in the RAM 1b as envelope attack rate data. Further, the performance speed data is multiplied by the amplitude envelope attack level shift data, the multiplication result is added to the amplitude envelope attack level data, and the addition result is corrected as amplitude envelope attack level data into the sound channel area in the RAM 1b. Is set. Further, the performance speed data is multiplied by the start address shift data, the result of the multiplication is added to the start address data, and the result of the addition is set as corrected amplitude start address data in the sound channel area in the RAM 1b.

[0075] The above-described shift table, as shown in FIG. 9, Starring Kanade velocity data horizontal axis average gate-on time AGO, the vertical axis is calculated as shift data (table output)
It is. Based on the performance speed data, a correction value of the shift data to be added to the tone control data is determined.

That is, in the process of step S613, if the average gate-on time AGO is shorter than the time T1, that is, if it is estimated that the performance speed up to immediately before and therefore the note-on performance speed this time is also high, the table The output "1" is multiplied by each shift data, and a shift for each tone control data is executed 100% according to the multiplication result.

As a result, as described above, in the case of a tone having a slowly rising waveform in the attack portion, large values are set as the amplitude envelope attack rate shift data and the amplitude envelope attack level shift data, and Is faster, the shift data is added to the amplitude envelope attack rate data and the amplitude envelope attack level data as it is, so that the rise of the amplitude envelope in the attack portion is faster and the maximum level in that portion is also larger. Data is generated.

On the other hand, in the case of a tone color containing a noise component in the attack portion, a large value is set as the start address shift data.
By adding the shift data to the start address data as it is, the start address of the tone waveform data to be read from the waveform data area in the ROM 1a is shifted so that the noise component of the attack portion is not generated.

Further, in the processing of step S613 described above,
If the average gate-on time AGO is longer than the time T2, that is, if it can be estimated that the playing speed up to immediately before is very slow and therefore the playing speed of the current note-on is also slow, the table output "0" is multiplied by each shift data, as a result,
No shift is performed for each tone control data.

As described above, the tone control data relating to the waveform readout position of the attack portion or the amplitude envelope characteristic is shifted in accordance with the performance speed, and
By setting the tone generation channel area b, a preferred tone corresponding to the currently recognized performance speed is emitted.

Finally, in step S614, the ROM 1a
From the data area corresponding to the currently set tone,
The tone control data other than the tone control data related to the waveform readout position or the amplitude envelope characteristic of the attack section is read out and set in the sounding channel area corresponding to the current sounding channel set in the sounding channel register in the RAM 1b. Is done.

Next, the note-off process in the process of S5 in the main flowchart will be described with reference to the flowchart of FIG. In step S1001,
It is determined whether a new note-off has occurred.

If a new note-off does not occur, the determination result of step S1001 is NO, and the note-off process is not executed. If a new note-on occurs, in step S1002, a note number (pitch data) set in the maximum 16 sounding channel areas of the RAM 1b is searched for a note number corresponding to the note-off note number.

In step S1003, the number of the sound channel obtained by the search is set in the sound channel register CH. Subsequently, step S100
At 4, it is determined whether or not the bit corresponding to the sounding channel number stored in the sounding channel register CH is "1" in the gate on time ignoring flag GOI.

If the decision result in the step S1004 is YES, the note-off this time is a note-off of a component sound other than the component note which was initially turned on in a component sound of one chord. And step S
The storage processing of the note-off time and the gate-on time in 1005 to S1009 is not executed, and the determination result in step S1004 becomes YES, and only the release processing in step S1011 is executed.

If the decision result in the step S1004 is NO, in a step S1005, the note corresponding to the current sounding channel stored in the sounding channel register from the current time value (current note-off time value) indicated by the time counter TC. The time value of the ON time register NOT (CH) is subtracted, and the subtraction result, the time difference ΔTC−NO
It is determined whether T (CH) ＣＨ is negative.

When the time difference {TC-NOT (CH)} is negative, the maximum value of the time counter TC is set as shown in FIG. 8B, as in the case of step S605 in FIG. A value “1000” which is one greater than “FFFF (H)”
0 (H) ”to the note-on time register NOT (CH)
The value obtained by adding the value of the time counter TC to the value obtained by subtracting the time value is the true time difference between the current note-off time and the corresponding note-on time. That is, if the decision result in the step S1005 is YES, a true time difference is calculated in a step S1006, and the calculated time difference is stored in the working register W as a gate-on time.

On the other hand, if the time difference (TC-NOT (CH)) is positive, the current time value (the current note-off time value) of the time counter TC is set to the note-on time register NO.
The value obtained by subtracting the time value of T (CH) is the true time difference between the current note-off time and the corresponding note-on time. That is, if the decision result in the step S1005 is NO, a true time difference is calculated in a step S1007, and the calculated time difference is stored in the working register W as a gate-on time.

Next, in step S1008, it is determined whether or not the gate-on time obtained in the working register W is less than 2000 msec in decimal notation. If the decision result in the step S1008 is YES, the gate-on time value of the working register W is rewritten to "2000 (decimal notation)", and the maximum allowable value of the gate-on time for calculating the performance speed is limited to 2000 msec or less.

If the decision result in the step S1008 is NO, the step S1009 is not executed. Step S1
At 010, the gate-on time obtained in the working register W and the current note-off time, which is the current time value of the time counter TC, are stored in the tone-generating channel register and correspond to the current tone-generation channel and the note-off time. It is stored in the time register GOF (CH).

In this way, every time a note-off occurs, the gate-on time and the note-off time corresponding to the note-off sounding channel are stored. These values are
It is used for the process of calculating the average value of the gate-on time in the process of step S612 in FIG. 6 described above.

Finally, in step S1011 a release process for attenuating the amplitude of the tone waveform data to mute it is executed. Next, the processing of the timer interrupt routine 2 in FIG. 11 will be described.

The interrupt control unit (not shown) in the CPU 1 interrupts the program corresponding to the main flow in FIG. 5 at each sampling timing, interrupts the processing of the program, and executes the timer interrupt routine 2 in FIG. Of the program corresponding to is started.

The timer interrupt routine 2 shown in FIG.
First, in step S1101,
Sound source processing is executed. FIG. 12 is a flowchart showing step S in FIG.
11 is a flowchart illustrating details of a sound source process of 1101.

First, in step S1201, the RAM 1b
The register R for accumulating musical tone waveform data is cleared.
Next, in steps S1202 to S1217, sound source processing is performed for each sounding channel. When the sound source processing for the 16th sounding channel is completed in step S1217, the tone R for the 16 sounding channels is stored in the tone waveform data accumulation register R. Musical sound waveform data obtained by accumulating waveform data is obtained.

The sound source processing for each sounding channel will be described. Each sounding channel area is provided in the RAM 1b of FIG.
Various control data for sound source processing based on the CM method are stored. Then, in the sound source processing corresponding to each sounding channel, a process of generating a musical tone is executed while using various control data set in each sounding channel area.

In the sound source processing for each sounding channel,
First, the start address data set in the self sounding channel area in the RAM 1b in the note-on processing of FIG. 6 in step S5 of the main flowchart of FIG.
Based on the pitch data set in the self-sounding channel area in M1b, waveform address data is calculated at each sampling timing, and the musical tone waveform data is read from the waveform data area in ROM 1a based on the waveform address data. Next, the read musical tone waveform data is multiplied by the amplitude envelope data. The amplitude envelope data is calculated based on the amplitude envelope plate data and the amplitude envelope level data set in the self-sounding channel area in the RAM 1b in the note-on processing of FIG. 6 in step S5 of the main flowchart of FIG. Multiplied by the data. The tone waveform data obtained in this manner is tone waveform data output corresponding to the self-sounding channel.

Here, as described above, the envelope plate data and the envelope level data for calculating the amplitude envelope data in the sound source processing for each tone generation channel are obtained in step S of the main flowchart of FIG.
In the note-on process of FIG. 6 in FIG. 6, when a key press is detected, rate data and level data based on the contents of the ROM 1a are set in a corresponding sounding channel area in the RAM 1b. In this case, as described above in the description of the note-on process in FIG. 6, the attack rate data and the attack level data are varied according to the performance speed data calculated based on the past average gate-on time. As a result, in the case of a tone whose waveform rises slowly in the attack portion, when the performance speed increases, the tone of the amplitude envelope in the attack portion rises quickly, and a tone having a large maximum level in that portion is generated. That is, even if a musical tone having a tone with a gradual rise is played early, a clear tone is obtained, and the desired tone is played.

As described above, the start address data for reading out the tone waveform data from the waveform data area of the ROM 1a in the sound source processing for each tone generation channel is also the same as the note-on data shown in FIG. 6 in step S5 of the main flowchart of FIG. In the process, when a key press is detected, start address data based on the contents of the ROM 1a is set in a corresponding sounding channel area in the RAM 1b. In this case, as described above in the description of the note-on processing in FIG. 6, the start address data is varied according to the performance speed data calculated based on the past average gate-on time. As a result, in the case of a tone containing a noise component in the attack portion, the start address of the tone waveform data to be read from the waveform data area in the ROM 1a is shifted so that the noise component in the attack portion is not generated when the playing speed increases. Thus, a musical tone from which a noise component has been removed is generated.

Each tone waveform data obtained by the sound source processing for each tone generation channel is sequentially accumulated in the tone waveform data accumulation register R. When the process of step S1217 in FIG. The musical tone waveform data obtained by accumulating the musical tone waveform data for the channels is obtained in the musical tone waveform data accumulation register R.

When the sound source processing in step S1101 of FIG. 11 shown in the flowchart of FIG. 12 is completed as described above, in the next step S1102, the tone waveform data obtained in the tone waveform data accumulation register R is replaced with the tone waveform data. The signal is latched by the latch 1c based on the sound source processing end signal COM.

The tone waveform data latched by the latch 1c in this manner is latched by the latch 1d by the interrupt signal INT synchronized with the sampling clock, and output to the D / A converter 1e.

In the embodiment described above, the characteristic of the rising portion of the musical tone, that is, the characteristic of the rising portion of the musical tone, that is,
Although the amplitude envelope characteristic and the waveform reading start position are controlled, the present invention is not limited to this.

That is, according to the present invention, when there is a case where a preferable tone is not produced depending on a kind of musical tone to be produced when the performance state is different, a new note-on is performed based on the evaluation result of the performance state in a past predetermined period. The present invention can be applied to all cases where the characteristics of a musical tone generated by an instruction are controlled.

[0105]

According to the present invention, the performance of a musical tone generated based on a new note-on instruction is controlled based on an evaluation result of a performance state in a predetermined period in the past, so that a currently recognized performance can be controlled. It is possible to generate a preferred musical tone corresponding to the state.

Specifically, for example, in the case of a tone containing a noise component in the rising portion, when the performance speed increases, the waveform generation start position of the musical tone is shifted so that the noise component in the rising portion is not generated. Thus, it is possible to generate a musical tone from which noise components have been removed.

For example, in the case of a tone whose waveform rises slowly, the amplitude envelope characteristic of the rising portion of the musical tone is changed so that the waveform rises quickly and its maximum level becomes large, so that the tone is sharpened. It is possible to produce rich musical tones.

[Brief description of the drawings]

FIG. 1 is an overall configuration diagram of an electronic musical instrument according to an embodiment.

FIG. 2 is a diagram showing an example of tone control data stored in a ROM.

FIG. 3 is a diagram illustrating a configuration of a register of a RAM.

FIG. 4 illustrates a configuration of a register of a RAM.

FIG. 5 is a main flowchart.

FIG. 6 is a flowchart of a note-on process in step S5 of the main flowchart.

FIG. 7 is a flowchart of a timer interrupt routine 1.

FIG. 8 is a diagram schematically showing a count state of a time counter TC.

FIG. 9 is a diagram showing a shift table.

FIG. 10 is a flowchart of a note-off process in step S5 of the main flowchart.

FIG. 11 is a flowchart of a timer interrupt routine 2.

FIG. 12 is a flowchart illustrating details of sound source processing.

[Explanation of symbols]

 1 CPU (Central Processing Unit) 1a ROM (Read Only Memory) 1b RAM (Random Access Memory) 1c, 1d Latch 1e D / A (Digital / Analog) Converter 2 Switch 4 LPF (Low Pass Filter) 5 Amplifier (Amplifier) ) 5 6 Speaker

Claims (3)

(57) [Claims] 1. A musical tone generated within a predetermined time in the past.
Performance state evaluation means for evaluating the performance state based on the average value of the time from toon to note off , and a new note-on instruction based on the performance state evaluation result by the performance state evaluation means. And a tone characteristic control means for controlling characteristics of a tone generated based on the tone control device. 2. The musical tone characteristic control means according to claim 1 , wherein
Based on the evaluation result of the performance state by the evaluation means, a new
Rise of musical tone generated based on note-on instruction
The tone control device according to claim 1, wherein an amplitude envelope characteristic of the section is changed . 3. The musical performance characteristic control means includes:
Based on the evaluation result of the performance state by the evaluation means, a new
Opening of waveform generation of musical tone generated based on note-on instruction
Shifting the start position, the musical tone control apparatus according to claim 1, wherein a.