JPH06149256A - Musical sound generator - Google Patents

Abstract

(57) [Abstract] [Purpose] Regarding a musical sound generator having a built-in effect adding device for adding a sound effect to a musical sound signal, the distribution level of each input musical sound signal when a plurality of effect adding processes are executed in parallel, The purpose is to be able to change according to performance data. [Configuration] Waveform data output from the tone signal generation circuit 106, each effector process #. 1 to # m, in the allocation level L ₁ ~L _m based on the average value of the velocity of key pressing sent from CPU103 Will be distributed. The output signals obtained by the respective effector processes of # 1 to #m are mixed with each other and also mixed with the original sound waveform data to which the effector process is not added, and then output to the D / A converter 108.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a tone generating device having a built-in effect adding device for adding a sound effect to a tone signal.

[0002]

2. Description of the Related Art Conventionally, an effect adding device has been used which adds a sound effect such as a reverb (reverberation), a chorus effect, or a time delay (time delay) to a musical sound played by an electronic musical instrument or the like.

Addition of these effects (effector processing)
Is often performed by itself, but when more complex and diverse effects are required in terms of performance expression, multiple types of effectors, such as a chorus effect and a reverb effect, are added. Often the process is performed.

In this case, there is known a technique capable of reflecting the intention of the performer on the effect of the effector by changing the characteristics of each effector process according to the performance data such as the velocity of each performance operation. ing.

[0005]

However, in the above-mentioned conventional example, there is a problem that sufficient performance expression cannot always be obtained only by changing the characteristics of the effector processing according to the velocity.

An object of the present invention is to determine the distribution level of each input musical tone signal when a plurality of effect adding processes are executed in parallel,
It is to be able to change according to performance data.

[0007]

The present invention has a tone signal generating means for generating waveform data of a tone signal based on input performance data.

Next, there is provided a plurality of effect adding processing executing means for executing the acoustic effect adding processing on the waveform data generated by the tone signal generating means. It also has distribution level control means for controlling the distribution level of the waveform data generated by the tone signal generation means to each effect addition processing execution means based on performance data such as velocity data. This means calculates, for example, a plurality of distribution level values based on velocity data, multiplies the waveform data generated by the tone signal generating means by each distribution level value described above, and each multiplication result is each effect addition processing execution means. To enter.

Then, it has an output means such as an adder which mixes the respective waveform data output from the plurality of effect addition processing execution means and outputs at least one output waveform data.

[0010]

The distribution level of the waveform data generated by the tone signal generation means to each effect addition processing execution means is controlled by the performance data such as velocity data obtained by the performance of the performer. Therefore, the performer can freely and in real time control what kind of effect addition processing is to be applied to the waveform data of the musical tone signal, based on his / her own playing state.

[0011]

Embodiments of the present invention will now be described in detail with reference to the drawings. The present embodiment is realized as an electronic keyboard instrument having an effect function.

The CPU 103 operates while using the RAM 105 as a work memory based on a control program stored in the ROM 104, scans the function keys 101 and the keyboard 102, captures the operation states of the function keys and keys, and controls the sound generation. The data is sent to the tone signal generation circuit 106 to control its sounding operation, and various settings for the effector processing are made to the DSP 107.

The DSP 107 is a tone signal generation circuit 106.
Waveform data of the generated musical sound is fetched, and a plurality of effector processes are executed on the waveform data as parallel time division processes.

The waveform data output from the DSP 107 is converted into an analog tone waveform signal by the D / A converter 108, the signal is shaped by the low pass filter (LPF) 109, then amplified by the amplifier 110, and the speaker 111.
Is emitted from.

Next, FIG. 2 is a block diagram of the DSP 107 of FIG. The respective units are connected by a bus 201, and the instruction decoder 203 executes a control program stored in the program memory 202, thereby executing multi-effector processing.

At this time, the multiplication operation in the effector processing is executed by the multiplier 206, and the addition / subtraction operation and the logical operation are executed by the arithmetic logic operation unit (ALU) 207. Also, various coefficients for determining the characteristics of the effector processing are
Various variables stored in the coefficient memory 204 and used in the effector processing are stored in the work memory 205.

The waveform data generated by the tone signal generation circuit 106 is transmitted through the waveform input interface 208 to the original tone buffer 4 and 4 described later in the work memory 205.
It is stored in one waveform buffer and multi-effector processing is executed.

Various control data set from the CPU 103 to the DSP 107 is stored in the CPU interface 20.
9 to the work memory 205. The output waveform data obtained as a result of the multi-effector processing is output from the waveform output interface 210 to the D / A converter 10 of FIG.
8 is output.

In the embodiment of the present invention having the above structure,
As will be described later, the DSP 107 can add up to four acoustic effects to the waveform data. That is, in the DSP 107, a maximum of four effector processes are executed as parallel time-division processes in each sampling cycle.

Next, the operation principle of the embodiment having the above-mentioned structure will be described. FIG. 3 is a diagram illustrating the principle of the embodiment of the present invention. In the embodiment of the present invention, the DSP 107 of FIG.
Can perform m kinds of effector processing as parallel time division processing on the waveform data output of the musical sound signal output from the musical sound signal generation circuit 106.

In this case, the above waveform data output is # 1.
Distribution level L based on the average value of the velocities at the time of key depression sent from the CPU 103 for each effector process of #m to #m
It is distributed at _{1 to} L _m . The total sum of L _{1 to} L _m is determined to be 1.

That is, in the embodiment of the present invention, the performer can freely control the degree of the effector processing depending on the strength of the velocity when the key is pressed. The performer can also select in advance the characteristics of how the distribution levels L _{1 to} L _m for each effector process are determined for each velocity value.

The output signals obtained by the respective effector processes of # 1 to #m are mixed with each other, and after being mixed with the original sound waveform data to which the effector process is not added, the D / A converter 108 of FIG. Is output to.

The specific operation of the embodiment of the present invention based on the above principle will be described below. First, FIG.
Is a CPU general flow showing the overall operation of the control operation executed by the CPU 103. This flow is realized as an operation in which the CPU 103 executes the control program stored in the ROM 104.

First, when the apparatus is powered on, in step S401, the contents of the RAM 105, the CPU 1
Contents of various registers in 03, tone signal generation circuit 106
The internal state of the DSP 107 and the internal state of the DSP 107 are initialized.

After that, each processing of steps S402 to S407 is repeatedly executed. In the function key acquisition process of step S402, a tone color setting key, an effector setting key,
The setting state of the function keys 101 such as the velocity characteristic selection key and the modulation key is scanned, and the setting state is stored in the RAM 105 or the like.

In the function key processing of step S403, processing corresponding to the setting of each function key 101, for example, processing for instructing the tone signal generation circuit 106 to change the timbre is executed. Further, an operation of generating a state table described later according to the setting state of the effector setting key which is the function key 101 and storing it in the RAM 105 is executed. Further, as a process particularly related to the present invention, the function key 101
The process of selecting the velocity conversion table is executed according to the selected state of the velocity characteristic selection key.

In the keyboard loading process of the next step S404, the key pressing state of the keyboard 102 is scanned, and the key pressing state is loaded into the RAM 105 or the like. Step S40
In the keyboard processing of No. 5, the RAM corresponding to the above key depression state is
Necessary pronunciation control data is set in the pronunciation control area provided in 105. In this case, as a process particularly related to the present invention, the average value of the velocities by the current key depression and the past five key depressions is calculated, and the velocity conversion table selected by the function key process of step S403 is obtained by the velocity. When accessed, the DSP 107 executes a process of calculating a distribution level (see FIG. 3) for effector processing, which will be described later. Details of this processing will be described later.

In the effect instruction process of step S406, the contents of the state table, which will be described later, generated in the function key process of step S403 according to the setting state of the effector setting key which is the function key 101, and step S4.
The distribution level calculated in the keyboard processing of 05 is DS
It is transferred to P107.

In the tone generation process of step S407, the tone generation control data set in the keyboard process of step S405 is transferred to the tone signal generation circuit 106, and a tone control instruction is given to the tone signal generation circuit 106. As a result, the tone signal generation circuit 106 executes tone generation control operations such as tone generation, muffling, and pitch change of tone waveform data.

The above is the processing of the CPU general flow. Here, as described above, the user is
The type and order of effector processing to be executed by can be arbitrarily set within a range of up to four. This setting can be performed by operating an effector setting key and a velocity characteristic selection key (not shown) which are the function keys 101.

When the user operates the effector setting key and the velocity characteristic selection key, which are the function keys 101, the operating states thereof are loaded into the RAM 105 by the function key loading process of step S402 of FIG. 4, and the function of step S403 is loaded. In the key processing, the state table is generated and the velocity conversion table is set.

First, FIG. 5 shows an example of the state table. The pointer value indicates the number of the effector process executed by the DSP 107, and is “00”, “01”, “10”, and “1”.
1 "is the first, second, and third in that order, respectively.
This corresponds to the effector processing executed at the fourth and fourth times.

The effector type based on the setting made by the user is set to each pointer value. In FIG. 5, effector types such as reverb 1, reverb 2, chorus, and delay are set. In addition, if there are n kinds of effector types that can be set, a code of the number of bits that can identify the n kinds of effector types is set in the effector type section of the state table.

When the user wants to execute only two effector processes, the user sets the effector type only for the pointer values "00" and "01" and sets the effector type for the other pointer values. Good if not. The effector process corresponding to the pointer value for which the effector type is not set is controlled so as not to be executed by the determination process of step S1102 of FIG. 11 described later.

The above-mentioned status table is obtained in step S of FIG.
It is transferred to the DSP 107 in the effect instruction processing of 406. When the above status table is set, the velocity conversion table is selected next.

This table is a table for calculating distribution levels (see FIG. 3) for maximum four effector processes set by the above-mentioned state table,
A plurality of tables are stored in advance in the ROM 104.

Now, in the status table, the pointer value "00"
When two effector types are set in correspondence with “01” and “01”, the velocity conversion table having the characteristic shown in FIG. 6 is selected based on the operation of the velocity characteristic setting key by the user. By using this table, distribution levels L ₁ and L ₂ (total sum is 1) for two effector processes corresponding to pointer values “00” and “01” in the keyboard process (to be described later) in step S405. ) Is calculated. A velocity conversion table similar to that of FIG. 6 for processing two effectors has a plurality of characteristics prepared in advance in the ROM 104 of FIG. 1, and the user can operate the velocity characteristics selection key from among these. , You can choose any table.

In the status table, the pointer value "0"
When three effector types are set corresponding to 0 ”,“ 01 ”, and“ 10 ”, the velocity conversion table of the characteristic shown in FIG. 7 is selected based on the operation of the velocity characteristic selection key by the user. By using this table, the pointer values "00", "01" and "10" in the keyboard processing of step S405.
Distribution level L for processing three effectors corresponding to
₁ , L ₂ and L ₃ (total sum is 1) are calculated. 3
A velocity conversion table similar to that of FIG. 7 for one effector process is prepared in advance in the ROM 104 of FIG. 1 for a plurality of characteristics. From among these, the user operates the velocity characteristic selection key to select an arbitrary characteristic. You can select the table.

Further, when four effector types are set in the state table in correspondence with the pointer values "00", "01", "10" and "11", based on the operation of the velocity characteristic selection key by the user. For example, a velocity conversion table having characteristics as shown in FIG. 8 is selected. By using this table, step S
In the keyboard processing of 405, pointer values “00” and “0”
Allocation levels L ₁ , L ₂ , L ₃ and L for processing four effectors corresponding to 1 ”,“ 10 ”and“ 11 ”.
₄ (total is 1) is calculated. Note that a velocity conversion table similar to that shown in FIG. 8 for processing four effectors has a plurality of characteristics prepared in advance in the ROM 104 of FIG.
From this, the user can select an arbitrary table by operating the velocity characteristic selection key.

Next, as described above, step S in FIG.
In the keyboard processing of 405, the average value of the velocities by the current key depression and the past five key depressions is calculated, and the velocity is used to access the selected velocity conversion table to calculate the distribution level. To be executed.

FIG. 9 is an operation flowchart relating to a part of the operation of the keyboard processing in step S405. First, in step S901, it is determined whether or not a key has been pressed.

If this determination is NO, no processing is performed. If the determination in step S901 is YES, in step S902, the velocity value corresponding to the depressed key captured in the keyboard capture process of step S404 of FIG. 4 is set to the value of the address pointer and the start address of the shift register area. Is added to the shift register area. In this case, the shift register area is a data area for, for example, 6 addresses provided on the RAM 105,
Each time a key press is detected, the velocity value corresponding to the key press is stored in the address indicated by the address pointer and sequentially updated within the above 6 addresses.

In step S903, the value of the address pointer is incremented by +1. When the value of the incremented address pointer exceeds "5", the determination in step S904 becomes YES and the value of the address pointer is "0". Returned to.

In this way, the six velocity values of the key that has been pressed most recently are stored in the shift register area. In step 906, the average value of the above six velocity values on the shift register area is calculated and used as the output velocity value. That is, each time a key depression operation occurs,
The moving average value of the six velocity values of the key that has been pressed most recently will be calculated.

In step S907, the velocity conversion table on the ROM 104 selected by the function key processing in step S403 is accessed by the output velocity value, and the distribution level L _i (1≤i≤4 (maximum), but the total sum is 1). ) Is calculated.

The distribution level L _i calculated in this way
3 is transferred from the CPU 103 to the work memory 205 in the DSP 107 in the effect instruction process of step S406 in FIG.

The process in which the DSP 107 receives the above-mentioned state table and distribution level is performed by the DSP described below.
For example, an interrupt process different from the general flow of 107 is performed.

Next, FIG. 10 is a diagram showing a DSP general flow showing the effector processing operation executed in the DSP 107. This flow is realized as an operation in which the instruction decoder 203 of FIG. 2 executes the control program stored in the program memory 202.

First, when the power of the apparatus is turned on, the pointer variable is reset in step S1001. This pointer variable is "00", "01", "1".
A 2-bit variable value of "0" or "11" is taken and provided in the work memory 205 of Fig. 2. Step S10
At 01, the value of the pointer variable is reset to "00".

Next, the processing of steps S1002 to S1007 is repeatedly executed in synchronization with the sampling cycle. First, in the voice input process of step S1002, one sample of waveform data output from the tone signal generation circuit 106 of FIG. 1 in the current sampling period is provided in the work memory 205 from the waveform input interface 208 of FIG. The four waveform buffers and the original sound buffer are loaded. For the waveform data stored in each of the four waveform buffers, DSP1
The waveform data stored in the original sound buffer after being subjected to the effector processing by 07 is processed in step S1007 described later.
In the audio output process of (1), the waveform data subjected to the effector process is mixed and output.

Next, the effector processing of step S1003 is repeated four times while the pointer variable is incremented in step S1004 and it is determined in step S1005 whether or not the value is "00". The effector processing is executed as parallel time-division processing.

FIG. 11 is a diagram showing an operation flow of the effector process of step S1003 of FIG. In this effector process, the effect instruction process of step S406 of FIG. 4 described above causes the CPU 103 of FIG.
The effector type of each effector process is determined by referring to the content of the state table (see FIG. 5) transferred to the work memory 205 in P107 based on the current value of the pointer variable.

Similarly, the CPU 103 causes the DSP 10 to perform the effect instruction processing in step S406 of FIG.
The distribution level L _i (1 ≦ i ≦ 4 (maximum)) transferred to the work memory 205 in 7 is referred to based on the current value of the pointer variable to determine the distribution level of each effector process. To be done.

That is, first, in step S1101, the code of the effector type corresponding to the current value of the pointer variable is read from the state table transferred from the CPU 103 of FIG. 1 to the work memory 205 of FIG. now,
When the pointer variable is initially “00”, the code corresponding to the effector type of reverb 1 is read in the example of the state table of FIG.

In step S1102, it is determined whether or not the effector type code is set in the state table in correspondence with the current pointer variable value. If the determination result is NO, the effector process is not executed. Note that FIG.
In this example, since the effector type is set for all pointer values, the determination in step S1102 will not be NO.

If the above determination result is YES, step S
In 1103, the waveform data is read from the first waveform buffer corresponding to the current pointer variable value “00” in the work memory 205, and in the next step S1104, the distribution level L ₁ corresponding to the pointer variable is converted into a waveform. The data is multiplied, and the result of the multiplication replaces the contents of the first waveform buffer.

Next, in step S1105, an effector processing subroutine corresponding to the effector type read from the state table in step S1101 is read and executed for the contents of the first waveform buffer described above. .

In this way, the waveform data of the musical tone signal output from the musical tone signal generation circuit 106 of FIG. 1 corresponds to the effector processing corresponding to the value "00" (reverb 1) of the pointer variable at the present number 1. , The distribution level L ₁ calculated from the velocity conversion table accessed by the moving average value of the velocities of the depressed keys.

When the effector processing of the effector type corresponding to the current pointer variable value "00" is completed by the processing of steps S1103 to S1105 above, the waveform data obtained as a result of the processing is stored in the work memory 205 in step S1106. Value of the current pointer variable in "0"
The first waveform buffer corresponding to "0" is overwritten, and the one-time effector processing of step S1003 in FIG. 10 is completed.

Next, in step S1004, the value of the pointer variable is incremented from "00" to "01" and the determination in step S1005 becomes NO, and then the effector process of step S1003 is executed again.

At this time, since the value of the pointer variable is "01", the code of the effector type corresponding to the current value "01" of the pointer variable is read from the state table in step S1101 of FIG. . In the example of FIG. 6, the code of the effector type of the reverb 2 is read.

In step S1103, the waveform data is read from the second waveform buffer corresponding to the current pointer variable value "01" in the work memory 205, and in the next step S1104, the read waveform is read. The data is multiplied by the distribution level L ₂ corresponding to the current pointer variable value “01”, and the result of the multiplication replaces the contents of the second waveform buffer.

Then, in step S1105, the above-mentioned 2
Step S1101 for the contents of the th waveform buffer
At, the effector processing subroutine corresponding to the effector type read from the state table is read and executed.

After the effector process of step S1003 of FIG. 10 is completed, the value of the pointer variable is incremented from "01" to "10" in step S1004, and the determination in step S1005 becomes NO, and then the step is restarted. The effector process of S1003 is executed.

At this time, since the value of the pointer variable is "10", the code of the effector type corresponding to the current value "10" of the pointer variable is read from the state table in step S1101 of FIG. . In the example of FIG. 5, the chorus effector type code is read.

In step S1103, the waveform data is read from the third waveform buffer corresponding to the current pointer variable value "10" in the work memory 205, and in the next step S1104, the read waveform is read. the data, the distribution level L ₃ which corresponds to the value "10" of the current pointer variable is multiplied, the contents of the third waveform buffer at the multiplication result is replaced.

Thereafter, in step S1105, the above-mentioned 3
Step S1101 for the contents of the th waveform buffer
At, the effector processing subroutine corresponding to the effector type read from the state table is read and executed.

After the effector processing in step S1003 of FIG. 10 is completed again, in step S1004,
After the value of the pointer variable is incremented from "10" to "11" and the determination in step S1005 is NO, the effector process in step S1003 is executed.

At this time, since the value of the pointer variable is "11", the code of the effector type corresponding to the current value "11" of the pointer variable is read from the state table in step S1101 of FIG. . In the example of FIG. 5, the code of the delay effector type is read.

In step S1103, the waveform data is read from the fourth waveform buffer corresponding to the current pointer variable value "11" in the work memory 205, and in the next step S1104, the read waveform is read. The data is multiplied by the distribution level L ₄ corresponding to the current pointer variable value “11”, and the result of the multiplication replaces the contents of the fourth waveform buffer.

Thereafter, in step S1105, the above-mentioned 4
Step S1101 for the contents of the th waveform buffer
Then, the effector processing subroutine corresponding to the effector type corresponding to the current pointer variable value "11" read from the state table is read and executed.

Finally, after the effector processing of step S1003 of FIG. 10 is completed, the value of the pointer variable is incremented in step S1004. However, since the pointer variable is a 2-bit variable value as described above, the pointer variable When the value "11" of is incremented, the value returns to "00".

In this way, the pointer variable is "00".
10 is reached, the determination in step S1005 of FIG. 10 becomes YES, and the four effector processes of steps S1003 to S1005 are ended.

In the timer processing of step S1006, every time an interrupt synchronized with the sampling period is input by the timer 211 of FIG. 2, the processing shifts to the audio output processing of the next step S1007. Then, in the audio output processing of step S1007, as a result of the effector processing of step S1003 performed four times, each output waveform data obtained in the four waveform buffers in the work memory 205 and the work memory 20.
The input waveform data previously stored in the original sound buffer in 5 is mixed, and the mixing result is the waveform output interface 2
It is output as output waveform data to the D / A converter 108 of FIG.

Here, steps S1002 to S1007
Since the processes up to are configured to be executed within one sampling cycle, as a result of the timer processing in step S1006, in the audio output processing in step S1007, the output waveform data is accurately synchronized with the sampling cycle in D / A. It will be output to the converter 108.

In the embodiment described above, as shown in FIG. 5, reverbs having different characteristics can be executed in parallel even with the same reverb, and depending on the setting, up to four types of the same reverb can be executed. It is possible to perform reverbs in parallel.

Further, in the above-mentioned embodiment, the distribution level for each effector process is determined by the moving average value of the velocity of the depressed key, but the present invention is not limited to this, and aftertouch is performed. The distribution level may be determined by performance data such as pitch data, pitch data, or modulation wheel data.

Further, in the above embodiment, the DSP 107
Is configured so that up to four effector processes can be executed as parallel time-division processes in each sampling cycle, but the present invention is not limited to this, and the DSP processing capacity has a margin. If so, it can be configured so that an arbitrary number of effector processes can be executed within a range of five or more.

[0080]

According to the present invention, the player can freely and in real time control what kind of effect addition processing is to be applied to the waveform data of the musical tone signal, based on his / her own playing state. It will be possible.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram of an embodiment of the present invention.

FIG. 2 is a configuration diagram of a DSP 107.

FIG. 3 is a diagram illustrating the principle of the embodiment of the present invention.

FIG. 4 is a diagram showing a CPU general flow.

FIG. 5 is a diagram showing an example of a state table.

FIG. 6 is a diagram (No. 1) showing a characteristic example of a velocity conversion table.

FIG. 7 is a diagram (No. 2) showing a characteristic example of a velocity conversion table.

FIG. 8 is a diagram showing a characteristic example of a velocity conversion table (No. 3).

FIG. 9 is an operation flowchart showing a part of the operation of keyboard processing.

FIG. 10 is an operation flowchart showing a DSP general flow.

FIG. 11 is an operation flowchart of effector processing.

[Explanation of symbols]

 101 Function Key 102 Keyboard 103 CPU 104 ROM 105 RAM 106 Musical Sound Signal Generation Circuit 107 DSP 108 D / A Converter 109 LPF 110 Amplifier 111 Speaker 201 Bus 202 Program Memory 203 Command Decoder 204 Coefficient Memory 205 Work Memory 206 Multiplier 207 Arithmetic Logic Arithmetic unit (ALU) 208 Waveform input interface 209 CPU interface 210 Waveform output interface 211 Timer

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Jun Hosoda 3-2-1 Sakaemachi, Hamura-shi, Tokyo Casio Computer Co., Ltd. Hamura Technical Center

Claims (2)

[Claims] 1. A musical tone signal generating means for generating waveform data of a musical tone signal based on inputted performance data, and a sound effect adding process for the waveform data generated by the musical tone signal generating means. A plurality of effect addition processing execution means; a distribution level control means for controlling distribution levels of the waveform data generated by the tone signal generation means to the respective effect addition processing execution means based on the performance data; An output unit that mixes each waveform data output from the effect addition processing execution unit and outputs at least one output waveform data. 2. The musical tone generating apparatus according to claim 1, wherein the performance data is velocity data.