JPH07168575A - Electronic musical instrument - Google Patents

Abstract

PURPOSE:To stabilize operation and to reduce scale relating to a space and a manufacturing cost by controlling a processing means by a program generating section and connecting a component corresponding to component information. CONSTITUTION:Component information constructing plural components being virtual modules is stored in a component file 900, component information stored in the component file 900 is selected by an operator with a selecting means, while, connecting state information defining wiring between components corresponding to component information is inputted by an input means. And programs of DSP and CPU are automatically generated by the processing means and a program generating section 300 based on the connecting state information, and a desired sound source algorithm is provided. Therefore, the operator himself of an electronic musical instrument can easily define an arbitrary sound source algorithm of the electronic musical instrument.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electronic musical instrument, and more particularly to an electronic musical instrument capable of creating an arbitrary algorithm regarding a sound source.

[0002]

2. Description of the Related Art Conventionally, an electronic musical instrument called a modular synthesizer has been known. The modules of this modular synthesizer are oscillator, filter,
It refers to a functionally lumped circuit configuration, such as an amplifier or envelope generator, and is a component of a synthesizer, so to speak.

An operator of a modular synthesizer could create an arbitrary sound source algorithm by wiring these modules using a patch cable. That is, the operator can construct a synthesizer having an arbitrary sound source algorithm by appropriately wiring the modules with patch cables, and based on such an arbitrary sound source algorithm, various musical sounds can be produced. It could be synthesized.

The above-mentioned sound source algorithm is a set of rules defining what kind of processing is to be performed and in what order in order to obtain a musical sound intended by the operator. The flow shows the order of processing.

[0005]

By the way, as described above, in the modular synthesizer, the modules having all independent circuit configurations are wired by the patch cables to construct the system. There is a high probability that contact failure and patch / cable connection will occur, resulting in a lack of operational stability.

Further, since each module has an independent circuit configuration, there is a problem that if all the modules that may be used are prepared, the scale becomes large in terms of space and cost.

On the other hand, in recent years, digitalization of synthesizers has progressed, and the sound source is a DSP (digital signal processor).
And a CPU (central processing unit), and the functions of the sound source algorithm and the sound source itself are realized by software that controls the DSP and the CPU.

However, in a synthesizer (digital synthesizer) using digital technology for such a sound source, in order to create an arbitrary sound source algorithm, D
An extremely advanced programming technique for controlling the SP and the CPU is required, and an operator who has not mastered such an advanced programming technique cannot freely create an arbitrary sound source algorithm.

The present invention has been made in view of the above-mentioned various problems of the prior art, and an object of the present invention is to make each component (module) of a sound source on a screen of a display device by software. Corresponding to a module of a type synthesizer, which is referred to as a "component" in the following description), and a conventional module is defined by programming the DSP or CPU based on this connection relationship. The simple operability when creating the sound source algorithm of the type synthesizer is realized by the digital sound source, and the operator can easily construct an arbitrary sound source algorithm in the digital sound source. An attempt is made to provide an electronic musical instrument that is stable and has a reduced space and cost. It is intended.

[0010]

In order to achieve the above object, the electronic musical instrument according to the present invention provides component information for constructing a plurality of components relating to musical sound generation in association with each of the above components. Storage means for respectively storing, selection means for selecting the component information stored in the storage means, and connection state information indicating an arbitrary connection state between the components corresponding to the component information selected by the selection means Input means, a programmable processing means for constructing the component and the connection state based on the component information and the connection state information, processing the input performance information and generating a musical tone, and the selecting means. The above component information selected by By automatically programming the processing means on the basis of the connection state information input by the input means, the processing means, the component corresponding to the component information selected by the selecting means, and the component And programming means for constructing the connection state indicated by the connection state information input by the input means.

[0011]

The storage means stores component information for constructing a component. The selection unit selects the component information stored in the storage unit, and the input unit inputs the connection state information indicating the connection state between the components corresponding to the selected component information. Then, the programming means controls the processing means to connect the components corresponding to the component information in the connection relation based on the connection state information.

That is, the component information for constructing a plurality of components as virtual modules is stored in the storage means, the operator selects the component information stored in the storage means by the selection means, and the input means. Then, the connection state information for defining the wiring between the components corresponding to the component information is input. Then, the processing means and the programming means automatically generate the programs of the DSP and the CPU based on the connection state information to realize an electronic musical instrument having a desired sound source algorithm.

Therefore, the operator of the electronic musical instrument can easily define any sound source algorithm of the electronic musical instrument.

The wiring between the components is CA
If a graphical user interface such as D is used, the operability is further improved.

[0015]

Embodiments of the electronic musical instrument according to the present invention will now be described in detail with reference to the accompanying drawings.

[Overall System Configuration] FIG. 1 shows a system configuration diagram of an electronic musical instrument according to an embodiment of the present invention. This electronic musical instrument has three processing units and these three processing units. It is composed of a data group used for processing.

The above-mentioned three processing units are DSP and CPU.
An algorithm for creating a sound source algorithm by defining a connection between a sound source unit 100 having hardware directly related to the generation of a musical tone and a component corresponding to a module and constructing wiring between the components. The generation unit 200 and the algorithm generated by the algorithm generation unit 200 are analyzed, and D of the sound source unit 100 is analyzed.
It is composed of a program generation unit 300 that generates a program for tone generation by the SP and the CPU.

A MIDI keyboard 400 for inputting performance information of the MIDI standard is connected to the sound source section 100, and further, an amplifier is used to emit a musical sound to a space based on a musical sound signal output from the sound source section 100. A sound system 500 including a speaker and the like is connected.

Further, in the algorithm creating section 200, the operator uses a display device 600 having a display screen such as a CRT and a keyboard 700 and a mouse 800 for inputting desired information to the algorithm creating section 200. The algorithm is created by operating the keyboard 700 and the mouse 800 while referring to the display screen of 600.

FIG. 2 shows a block diagram of the tone generator section 100, which includes a MIDI interface (I / F).
The system program R is connected to the bus 106 of the CPU 104 which is connected to the MIDI keyboard 400 via 102.
Communication interface (I / F) 1 between OM 108, work RAM 110, and a host computer described later
12, a PCM waveform reading device 114, three DSPs 116 connected in series, and an output selector 118 are connected. Further, a waveform memory 120 is connected to the PCM waveform reading device 114, and three DSPs 116 are connected. Are connected to the delay memories 122, respectively, and the output selector 1
A D / A converter 124 is connected to 18.

As shown in FIG. 2, the tone generator section 100 has an independent hardware structure, but the algorithm creating section 200 and the program creating section 300 are referred to as a workstation computer (hereinafter referred to as a host computer). The host computer has a communication interface connected to the communication interface 112 of the tone generator 100, and the host computer and the tone generator 100 are connected via these communication interfaces.) It is realized as an application program of. Therefore, the host computer includes the display device 600, the keyboard 700, and the mouse 800.

The configuration is not limited to the above, and the algorithm creating section 200 and the program creating section 30 are not limited thereto.
0 and 0 may have independent hardware configurations, or the tone generator 100, the algorithm generator 200, and the program generator 300 may have an integrated hardware configuration.

As the data group, a component file 900 storing component information for constructing a component corresponding to a module, and a PCM waveform data file 1000 storing musical tone waveform data for generating musical tones. Is set. These component file 900 and PCM waveform data file 1000 are also set on the host computer.

Here, the component means a component of the synthesizer corresponding to the module as described above, and in the present embodiment, TVF (time-varying filter), TVA (time-varying amplifier), EG (envelope. Generator), LFO (Low Frequency Oscillator), Chor
Us (chorus) and the like are defined.

The component file 900 is necessary for realizing the function of such a component.
It is a set of manual information in which the programs of the DSP 116 and the CPU 104 of the sound source unit 100, information of input / output terminals, specifications, etc. are recorded, and is collected in one file for each component. A detailed description of the component file 900 will be given later.

The components stored in the component file 900 will be described later in detail with reference to FIGS. 12 to 61.

The algorithm creating section 200 and the program creating section 300 read the information necessary for processing, such as the component file selected from the component file 900, and perform the necessary work.

On the other hand, the PCM waveform data file 100
0 is composed of the following data necessary for the PCM waveform reading device 114 of the tone generator 100 to generate a musical tone signal.

[Read Address / Data] Stores address information such as start and end of waveform reading and loop reading section, and flag information indicating whether or not to perform loop reading.

[Waveform data table] PC of musical tone waveform
Store M data.

The component file 900 and the PCM waveform data file 1000 are stored in the sound source unit 1.
00, the algorithm creating unit 200, and the program creating unit 300 are independent, and additional changes can be freely made. The flexibility of this additional modification enhances the flexibility and expandability of the entire system.

[Sound Source Unit 100] Among the constituent elements of the sound source unit 100 shown in FIG. 2, it is the PCM waveform reading device 114 and the DSP 116 that actually generate / stop and process the tone signal. . In addition, in this embodiment, three DSPs 116 are connected in series.

The PCM waveform reading device 114 is a CPU
The waveform data is read from the waveform memory 120 based on the pitch data and the read address data given from 104, and is output to the DSP 116 as an audio signal. Waveform reading is performed by time division, and independent audio signals of 16 channels can be output to the DSP 116.

The DSP 116 is a device capable of performing a convolution operation at high speed, and is capable of forming various filters for processing musical tone signals. The output of the PCM waveform reading device 114 is DSP1
16 is input.

The DSP 116 comprises an arithmetic unit, a program memory, a register memory, and a coefficient memory, and the delay memory 122 is connected to the DSP 116.

The program memory of the DSP 116 stores rewritable and transferred programs, and the arithmetic unit sequentially performs arithmetic processing according to the programs stored in the program memory. Further, the register memory is used to temporarily store a value in the middle of the calculation by the calculation device, but since it is configured as a ring buffer, the signal can be delayed. The coefficient memory is a memory that stores the coefficient of the convolution operation.

Further, the delay memory 122 has a large ring
A buffer used for long signal delays.

Delaying a signal by using a ring buffer is shown in, for example, Japanese Patent Publication No. 63-32393.

Each of the plurality of DSPs 116 connected in series has an output for D / A conversion, and the output selector 118 selects one of the outputs for D / A conversion of the DSP 116 and outputs the D / A conversion. It will be input to the A converter 124.

Further, the CPU 104, based on the system program stored in the system program ROM 108, the PCM waveform reading device 114 and the DSP 11
6 and the output selector 118 are controlled.

In the system program ROM 108,
As described above, the system program for controlling the operation of the CPU 104 is stored, and when the CPU 104 is activated, the system program stored in the system program ROM 108 is executed.
This system program uses a real-time monitor, and a plurality of tasks (programs) described below operate simultaneously in parallel in a time-sharing manner.

(1) MIDI Task The MIDI task is activated by a MIDI reception interrupt from the MIDI interface 102, analyzes received data, and performs processing corresponding to the received data.

When the received data is note-on or note-off, the assigner task is activated,
When the received data is a control change, vendor, after-touch, etc., the data is stored in the MIDI control buffer of work RAM 110 described later.

(2) Assigner task The assigner task is activated by the MIDI task in response to the occurrence of a note-on event and a note-off event.

First, note-on from MIDI task
When activated in response to an event, the sound generation channel allocation process of the PCM waveform reading device 114 is performed. The number of assigned sound generation channels is changed in Channels of the command list window 606 described later.

After allocating the tone generation channel, the note number and on-velocity corresponding to the note-on are stored in the corresponding channel buffer.

Next, the waveform reading of the assigned tone generation channel of the PCM waveform reading device 114 is stopped, the read address data is given, and the waveform reading is ready to start.

The tone generation control program is called with the channel number assigned last as an argument.

On the other hand, note off from MIDI task
When activated in response to an event, it refers to the channel buffer among the assigned channels, finds the sound channel to which the same note number as the note-off event is assigned, and releases the key. Identify the pronunciation channel that should be processed. The status flag of the tone generation channel is released and the off velocity is written to the buffer. Also, the key release control program is called with the channel number as an argument.

(3) Periodic Service Task The periodic service task is activated by a timer built in the CPU 104 periodically to control the tone generator 100, and is generated by the program generator 300. The control program is called. Note that the term service is a component that performs parameter calculation at regular time intervals and that is particularly time-varying (such as lfo and e described later).
nv, etc.) is always calculated.

(4) DSP Program Writing Task The DSP program writing task is activated when the CPU 104 receives the DSP program created by the program generator 300 from the host computer, and the DSP program is loaded into the program memory of the DSP 116. Write in.

(5) CPU control program write task The CPU control program write task is a CPU control program created by the program generation section 300 (three programs called an assigner task and a regular service task, a tone generation control program and a key release control). Programs and regular control programs) are started when the CPU 104 receives them from the host computer, and the CPU control programs are written in the control program area set in the work RAM 110.

Further, when the CPU 104 receives data from a host computer by changing processing of a fader (fader component: an operator can arbitrarily set a value manually), which is one of the components described later, the fader Write to the work memory for the component (corresponding buffer provided in the control program work area of the work RAM 110 described later).

(6) Waveform memory writing task The CPU 104 is activated when the waveform data is received from the host computer, and the received waveform data is written in the waveform memory 120. At the same time, the read address data is also sent from the host computer.
Store in buffer. Although the waveform memory 120 is not directly connected to the CPU 104 via the bus 106, the waveform data can be written via the PCM waveform reading device 114.

Next, the structure of the work RAM 110 will be described. The work RAM 110 is divided into three areas: a system work area, a control program area, and a control program work area.

(1) System work area The system work area is an area used as a work memory by a system program that executes each task described above, and holds the following data.

[Channel number buffer] The setting is changed in Channels of the command list window 606. It is used as the number of sound generation channels used in the PCM waveform reading device 114 and the number of replicas of the algorithm using the polyphonic window. The maximum value is
It is "16".

[Channel Buffer] The following data is held for each sound generation channel. (1) Note number (2) On velocity (3) Off velocity

[Read Address Buffer] Holds the read address for PCM waveform read.

[MIDI control buffer] Vendor information, control change information, channel
It holds the latest values of after touch information and polyphonic after touch information, respectively.

[Tense Data Buffer] Stores tense data for each component.

(2) Control Program Area The control program area is an area used for storing the CPU control program automatically generated by the program generation unit 300 by the processing described later.

(3) Control Program Work Area The control program work area is an area used as a work memory by the CPU control program automatically generated by the program generation unit 300 by the processing described later.

[Algorithm Creating Unit 200] The algorithm creating unit 200 is realized as an application program on the host computer as described above, and is used by the operator to create sound source algorithm data. This application program
Like the CAD software, a display device 600, a keyboard 700 and a mouse 800 are used to graphically define algorithms.

FIG. 3 shows the display state of the screen of the display device 600, and the components are displayed as rectangular icons in the algorithm drawing area 602 of the algorithm creation window 601 on the screen of the display device 600 (see FIG. 3 shows components such as Osc and Pan).

The operator selects a desired component from the component list area 604 displaying the name of the component, arranges the icon of the component on the screen as appropriate, and connects the arranged icons with a line. This line represents the flow of signals between the components for actually passing data through the program.

Reference numeral 606 is a command list window, reference numeral 608 is a help window showing information on the designated component, and reference numeral 610 is a polyphonic window used in a polyphonic process described later.

The display state of the help window 608 shown in FIG. 3 shows a case where a help instruction is given in a state where Pan is selected as a component. The help window 608 displays the function of each pin arranged in the component. This help instruction is input using the keyboard 700.

Further, in this embodiment, the DSP 116
Since three DSPs are provided, three algorithm creation windows 601 (DSP1, D1,
SP2 and DSP3).

The icon of the component is rectangular as described above, and the pins are arranged around it.
This pin indicates a signal input / output terminal.By connecting the pin of an icon and the pin of another icon, a component and another component are connected, and this work is repeated. Define the algorithm.

By the way, the algorithm indicates the flow of signals between the components, and the signals flowing between the components are classified into two types of signals.

One of these two types of signals is a signal called an audio signal, and finally the D / A of the sound source section 100.
It is a signal output as a tone signal via the converter 124. A component for generating and processing this audio signal will be referred to as an audio component.

The other signal is a signal called a control signal, which is a signal for controlling various parameters of the audio component. A component for generating and processing this control signal will be referred to as a control component.

The voice component is the DSP 116.
And the control component is the CPU 104.
Composed by.

The pins are also classified according to the type of input / output signals and the arranged side, that is, the input side of the signal or the output side of the signal, and this is called the pin attribute. Therefore, for the pin, the audio signal input, the audio signal output, the control signal input, and the control signal output are set as the attributes, and the trigger input and the trigger output regarding the trigger pin described later are also set as the attributes of the pin. It is set.

The attributes of these pins are preset in the component file 900 as pin attributes PinAttr.

By the way, the position of the pin arranged on the icon is determined by the attribute of the pin. That is, the voice input pin is arranged on the left side of the icon and the voice output pin is arranged on the right side of the icon. The control input pin and the trigger input pin are arranged on the upper side of the icon, and the control output pin and the trigger output pin are arranged on the lower side of the icon.

By adopting the pin arrangement as described above, when each pin is connected by a line, it seems that the audio signal flows in the horizontal direction and the other signals flow in the vertical direction. Therefore, the visibility of the algorithm is improved.

The component file 900 is shown in FIG.
As shown in, it is composed of the following data.

[Number of Pins PinNum] Indicates the number of pins of the component.

[Pin Information PinInfo] The following information regarding pins is stored for the number of pins. (1) Pin name PinName Indicates the name of the pin. (2) Pin attribute PinAttr Indicates the type of signal handled by the pin and the input / output direction. (3) Pin symbol PinSym The pin symbol PinSym is set only for the voice input / output pin and represents the symbol name of the corresponding input / output work RAM. (4) Pin tense PinTense Pin tense PinTense is set only on the input pin of the control signal, and indicates which control program to be described later implements the other components connected to the control pin.

[DSP Program Template Dsp
Tpl] Store the template of the DSP program.

[DSP Symbol Information DspSym] DS
Information for allocating the work memory (symbol) used in the P program template DspTpl is stored, and the following information is stored corresponding to all the symbols used in the template. . (1) Symbol name DSNName represents a symbol name. (2) Memory area DSArea Indicates the memory area to which work memory is allocated. There are three types: register memory, coefficient memory, and delay memory. (3) Reference address DSRef Indicates the position in the DSP program template DSPTpl that refers to the symbol. (4) Initial value This represents the initial value of the DSVal symbol.

[DSP Program Memory Usage SzP
rg] Indicates the size of the program memory used by this component.

[DSP Register Memory Usage SzRe
g] Represents the size of the register memory used by this component.

[DSP Delay Memory Usage Amount SzDel] This represents the size of the delay memory used by this component.

[Control Program Template Ctrl
[Tpl] A plurality of control templates can be stored,
Each template consists of the following data. (1) Template Tpl Store the template body. (2) Template tense TplSense stores tense data of the template.

[Control Symbol Information CtrlSym] Stores information for allocating the work memory (symbol) used in the template of the control program, and corresponds to all the symbols used in the template. Then, the following information is stored. (1) Symbol name CSName Indicates a symbol name. (2) Reference address This indicates the position where the symbol is referenced in the CSRef template. (3) Initial value CSVal Indicates the initial value of the symbol.

[Help Information Help] Stores a document describing the specifications and usage of the component.

As shown in the attached drawings, the icon of the component and the pattern of the pin are respectively determined by the attributes of the pin arranged in the component,
It is formed so that it can be easily identified on the screen of the display device 600.

That is, in the component file 900, information such as the number of pins arranged in the component, the name and attribute of each arranged pin is stored in advance for each component. Therefore, the algorithm creation unit 200 uses the component file 900.
An icon as described above is formed in the algorithm drawing area 602 from the above information stored in.

The drawing data of the component is created by combining pixels based on a quadrangle as shown in FIG. 5 (a).

The size of the quadrangle is determined by the number of pins to be arranged around the side and the length of the component name displayed inside. That is, the maximum of the following three expressions is the horizontal side length lh. (1) (pci + pti + 1) × pdh (2) (pco + pto + 1) × pdh (3) lnh + 2mnh where pci: number of control signal input pins pti: number of trigger signal input pins pco: number of control signal output pins pto: number of trigger signal output pins pdh: Pin spacing in the horizontal direction lnh: Length of component name mnh: Margin in the horizontal direction when displaying the component name.

Similarly, the vertical side length lv takes the maximum value of the following three expressions. (1) (psi + 1) × pdv (2) (pso + 1) × pdv (3) lnv + 2mnv where psi is the number of audio signal input pins pso is the number of audio signal output pins pdv: Pin spacing in the vertical direction lnv: Vertical direction of the component name Length mnv: Vertical margin when displaying component name.

Of the above equations, the number of pins pci, pc
o, pti, pto, psi, and pso can be known by counting the pin information PinInfo of the component file 900 while checking the pin attribute PinAttr.

The size of the display area for the component name is determined by the number of characters in the component name and the size of the font used.

Small squares indicating pins are left-aligned and upper-aligned with pd on each side of the rectangle representing components.
They are arranged at intervals of h or pdv. At that time, a black square (FIG. 5B) is used for the control signal pin and the audio signal pin, and a white square (FIG. 5C) is used for the trigger signal pin.

The component name is printed inside the rectangle with a margin of mnh and mnl from the upper left.

Further, in the quadrangle, except for the area where the component name is displayed, it is filled with a pattern corresponding to the characteristics of the component. A component having only audio signal pins, that is, an audio component is shown in FIG.
The component filled with the pattern of (d) and provided with only the control signal pin and the trigger signal pin, that is, the control component is filled with the pattern of FIG. 5 (e) and provided with the control signal pin, the trigger signal pin, and the audio signal pin. The component is filled with the pattern shown in FIG.

As described above, the names of the components stored in the component file 900 are displayed in the component list area 604 on the screen of the display device 600, and the components displayed in the component list area 604 are displayed. Are freely selected and placed in the algorithm drawing area 602, and the icons of the placed components are respectively wired. These operations are performed by operating the mouse 800.

That is, the instruction by the mouse pointer of the mouse 800 is used to give selection / arrangement of components, wiring of pins, and various other instructions to the program. The keyboard 700 is a fader
It is used for setting parameter names of components and setting file names when saving data.

When wiring between the icons of the components, the pin to be wired is first designated by the mouse 800. When the mouse 800 is clicked, the unwired pin closest to the mouse pointer will be designated.
Then, when the mouse 800 is moved to the position of the pin of the wiring partner and clicked again, the pin of the wiring partner is specified. Also in this case, an unwired pin closest to the mouse pointer and corresponding to input / output by the same type of signal is automatically selected and wired. In this way, since the pin of the wiring destination is automatically selected, wiring can be surely performed easily.

The distance between the mouse pointer and the pin is obtained from each coordinate, and the pin coordinate is the sum of the component coordinate and the relative coordinate of the pin in the component. Also, the relative coordinates of the pins are determined when the icon for the component is created.

The wiring of the pins described above is performed in a one-to-one relationship between pins.

Further, the wiring data when the pins are wired to each other is stored in the program in the following format, and a collection of such wiring data structures is the sound source algorithm data.

Data name Example of recorded contents Data description COMPO #: 0 ・ ・ ・ Serial number of the component being used C_NAME: Osc ・ ・ ・ Component name C_POSX: 353 ・ ・ ・ Component position coordinate (X coordinate) C_POSY : 1187 ... Arrangement position coordinates (Y coordinate) of the components The arrangement position coordinates of C_POSX and C_POSY are coordinates in the algorithm drawing area 602. C_PINS: 3 ・ ・ ・ Number of pins PIN #: 0 ・ ・ ・ Pin serial number in the component P_ATTR: 1 ・ ・ ・ Pin attribute number P_NAME: output ・ ・ ・ Pin name P_LNKC: 1 ・ ・ ・COMPO # of wiring destination P_LNKP: 0 ... PIN # of component of wiring destination PIN # PIN #: 1 ... Pin number in the component P_ATTR: 2 ... PIN attribute number ... Follows the same format for the number of pins. That is, first, information about components (COMPO # to C_PINS)
, Followed by as many pin information as C_PINS.

The serial numbers of the components are given in order when the components are placed in the algorithm drawing area 602. The pin number is the order of the pin information stored in the component file 900.

The pin attribute numbers are set as follows. 0 ... Voice signal input 1 ... Voice signal output 2 ... Control signal input 3 ... Control signal output 4 ... Trigger input 5 ... Trigger output

Then, when the pins are wired, P_LNK
Information on the wiring destination pin that is the wiring partner is written in C and P_LNKP. Further, the information of the wiring source pin is also written in the wiring destination P_LNKC and P_LNKP. It should be noted that when the pins are not wired, P_LN
Both KC and P_LNKP are set to "-1".

That is, when the operator selects a component and places it in the algorithm drawing area 602, the structure of the component is added to the algorithm data. That is, for each rendered component,
The above-mentioned wiring data is created, and these wiring data are integrated to form the sound source algorithm data.

That is, when the component is placed in the algorithm drawing area 602, P_LNKC and P_L
NKP is initialized to "-1", and when pins are wired, the information is rewritten to information corresponding to the wiring. The data created by the above-described operation is output to the program generation unit 300 as sound source algorithm data.

When a component is deleted, the structure of the component is deleted from the algorithm data. At this time, the line connected to the component is also erased from the algorithm drawing area 602, and "-1" is written to the pins P_LNKC and P_LNKP of the partner component to which the line is connected. It should be noted that even if the connection is deleted or the destination pin of the connection is changed, the contents of P_LNKC and P_LNKP corresponding thereto are changed.

From such sound source algorithm data, it can be known which pin of which component has which attribute and which pin of which component has which attribute. Therefore, it is possible to trace the flow of the audio signal and the control signal based on the information.

Further, since the resources of the DSP 116 of the tone generator section 100 (program memory, register memory, delay memory) are limited, the algorithm creating section 200
Always calculates what percentage of the resources of the DSP 116 the created algorithm uses, and displays it in the used resource display area 612 as shown in FIG. 3, for example.

That is, "P-RAM" (program memory) is "7.0%", "I-RAM" (register memory) is "1.6%", and "E-RAM" (delay memory). Is displayed as "0.0%". And the usage is 10
If it exceeds 0%, an error message is displayed to call attention and the operator is informed accordingly.

In the component file 900, the usage amount of each resource for each component is stored in advance, and the usage amount of the entire algorithm is obtained using the stored contents.

As described above, the algorithm creating unit 200 is the only user interface in the entire system configuration according to this embodiment, and in addition to creating the algorithm, the algorithm creating unit 200 By operating the command list window 606 (FIG. 6), the following process is instructed.

[Compile] Compile processing. The compilation here means that the algorithm data created by the algorithm creating unit 200 is converted into the program creating unit 30.
It refers to a series of processes for sending a program to 0 to generate a program and transferring the generated program to the tone generator 100.

[Save] The algorithm data created by the algorithm creating section 200 is saved in a file.

The algorithm data stored in the [Load] file is read into the algorithm creating section 200.

[Quit] This application program is terminated.

[Channels] The number of sounding channels is set. The maximum value that can be set is "16", which is the same as the number of channels of the PCM waveform generator.

When a complicated algorithm is created, the program to be generated also becomes large, and the program size does not fit within the limited program size, or the reaction speed of the musical tone control to the performance is lowered. Therefore, the program size and reaction speed can be adjusted by changing the number of sound generation channels.

[Waveform] Prepared PC
When a list of M waveform data files 1000 is displayed and an operator selects a desired waveform data file from the list, the waveform data table and read address data are read from the waveform data file. , To the sound source unit 100.

[DSP1] [DSP2] [DSP3] Algorithm creation windows 601 (DSP1, DSP2 and DSP) corresponding to the three DSPs 116, respectively.
3) is displayed on the display device 600.

A [Polyphonic] polyphonic window 610 is displayed on the display device 600.

(Regarding Processing of Plural DSP Chips) In this embodiment, plural DSPs are used as the DSP 116.
Since the SP chip is used, the following special processing is required. That is, in this embodiment, the sound source unit 100
In the case where the three DSPs 116 are provided, the processing in such a case will be described below. Note that the algorithm that can be executed by one DSP has a limit of DSP capability, so multiple DSs can be used.
By using P-chips, more sophisticated algorithms can be created.

That is, in the present embodiment, as shown in FIG. 3, the program generation section 300 is adapted to create one DSP 116 algorithm in one algorithm creation window 601.

Therefore, when using a plurality of DSP chips, an algorithm is created in each algorithm creation window 601 for each DSP chip, and these are connected together.

As shown in FIG. 2, the DSP 116 is connected in series and can simultaneously transmit a plurality of audio signals. This delivery is performed as shown in FIG.
Allows programming using InN, SerOutN components. In addition, SerIn
N in N and SerOutN is the number of the audio signal channel between the DSPs, from SerOut to Se of the same number.
An audio signal is sent to rIn.

Similarly to the above, as for the control signal, as shown in FIG. 7, the components ctrlInN and ctrlOutN are used to exchange signals between windows.

(Regarding Polyphonic Processing) In the polyphonic window 610, the algorithm for one note created here can be automatically created for the number of simultaneous sounds.

Normal algorithm creation window 601
In the algorithm drawing area 602 of 1., when an attempt is made to create a polyphonic sound source, a plurality of same algorithms must be drawn. Since the amount of work increases significantly and the efficiency is low, a window called a polyphonic window 610 is set separately from the algorithm drawing area 602 of the normal algorithm creation window 601, and the algorithm created in this window is set. It is automatically duplicated according to the number of generated sound channels.

The audio signal and the control signal processed by the algorithm created in the polyphonic window 610 are temporarily stored in the normal algorithm drawing area 602.
And then mixed there, D / A converter 12
4 is output. If necessary, some processing may be further performed in the normal algorithm drawing area 602.

FIG. 8 shows the polyphonic window 61.
It is an example of algorithm creation using 0.

In the polyphonic window 610,
The audio signal generated by the oscillator Osc is divided into two audio signal series by the Pan component, and the two series of audio signals are added together in a normal algorithm creation window. As a result, it is possible to realize an algorithm for performing stereo output with different sound image localization for each sounding channel.

The ToneOut component is a virtual component prepared for algorithm creation in order to pass an audio signal from the polyphonic window 610 to the normal algorithm creation window 601, such as "-A" or "-B". The identifier can handle up to four series of audio signals from the algorithm corresponding to one sound channel.

NoteIn component is Tone
It is a component that forms a pair with the Out component, and is a component for extracting the audio signal from the polyphonic window 610 on the side of the normal algorithm creation window 601. Identifiers such as "-0" and "-1" correspond to channel numbers, and the four output pins correspond to four series of ToneOut components (see "PinInfo 1" window).

Further, asgnN is a virtual component for creating an algorithm of an assigner component, and is used in the polyphonic window 610 whose channel number cannot be specified.

When the compilation is started by operating the Compile button, the algorithm creating section 200
Prior to the analysis of normal algorithm data, a process of copying the algorithm created in the polyphonic window 610 according to the set number of sounding channels is performed.

The algorithm data created in the polyphonic window 610 is stored in the same format as the algorithm data of the normal algorithm creation routine described above.

In the duplication processing, the algorithm data created in the polyphonic window 610 is converted into the algorithm data in the normal algorithm creation window 601.
This is performed by additionally copying the set sound generation channels to the data several times. At that time, the following modifications are made to the algorithm data.

The component number COMPO # in the algorithm data created in the polyphonic window 610 renumbers the component numbers used in the addition destination algorithm data.

If the asgnN component is used, its C_NAME data is the name of a normal assigner component asgnN corresponding to the channel number.
Replace with m (m is the channel number)

ToneOut-A, ToneOut-
The connection information is modified on the algorithm data so that the input signal to B is output from Note In-ch (ch is a channel number). This process is performed in the following procedure.

(1) The connection data P_LNKC and P_LNKP of the input pin of ToneOut-A are checked to obtain its component number COMPO # and pin number PIN #.

(2) Corresponding component number COMPO
The data P_LNKC and P_LNKP indicating the connection destination of the relevant pin number PIN # of the component #
eIn-ch output pin tone-A connection destination data P
Replace with _LNKC and P_LNKP.

(3) Similarly, the connection destination data P_LNKC and P_LNKP of the pin of the connection destination component of the output pin tone-A of the NoteIn-ch is set to Ton.
It is replaced with the connection destination data P_LNKC and P_LNKP of the input pin of eOut-A.

(4) The same processing is performed for ToneOut-B.

(5) After that, Ton which has finished the correction process
Delete the eOut and NoteIn component algorithm data.

The above-mentioned rewriting of the algorithm data for one channel is repeated for the number of tone generation channels set to be added. As a result, the algorithm created in the polyphonic window 610 can be formed corresponding to the number of channels for simultaneous sound generation.

The resource usage of the DSP 116 is also calculated by multiplying the number of sound generation channels for the components arranged in the polyphonic window 610, and calculating the first D
It is displayed in addition to the usage amount of SP116.

Similarly, the control signal is changed from the polyphonic window 610 to the normal algorithm creation window 6
01 algorithm drawing area 602, and a virtual component for that purpose is prepared.

That is, the polyphonic window 610.
CtrlPolyOutN (not shown) of the normal algorithm drawing area 602 of ctrlPolyIn
The control signal is transferred to N (not shown). Note that N
Is the channel number.

[Program Generation Unit 300] The program generation unit 300 analyzes the sound source algorithm data created by the algorithm creation unit 200, and creates the programs for the DSP 116 and the CPU 104.

Then, the DSP program generated by the program generation unit 300 outputs the voice component to the DSP.
It is intended to be configured on 116, and the characteristics of each component can be changed by changing the coefficient corresponding to the parameter.

Further, the CPU control program generated by the program generation unit 300 sets the control component as CP.
The control signal for controlling the coefficient corresponding to the parameter of the audio component is obtained from various performance information, and the DSP 11
6 controls the audio component.

The component file 900 stores a "template" which is a model of a DSP program and a CPU control program for realizing the function of the component, and "symbol information" corresponding to the template. The "template" mentioned here is a relocatable program routine, and the work memory used by the routine is not yet allocated.

The "symbol information" is information on the work memory used by this "template".
The symbol name for identifying the memory, the memory area to be allocated, the reference address indicating the position of the work memory in the template, and the initial value of the work memory are stored for each work memory used ( The control symbol information does not have the memory area information.
This is because the control program work area has only one control program work area. ).

The program generator 300 connects the template of the component according to the algorithm data and allocates the work memory to a predetermined memory area of the memory actually used according to the symbol information of the template. Further, the memory access instruction of the connected template is changed based on the reference address data so as to access the work memory allocated by the template. Then, if the initial value of the work memory is set, the initial value is stored in the assigned work memory.

When the above processing is completed for all components of the algorithm data, the linked templates are transferred to the program memory.

That is, the pin wiring is realized by accessing the work memory in the program. That is, the output pins will be written to the work memory and the input pins will be replaced with instructions to read the data from the work memory. By making the write and read work memories the same, it is possible to transfer data from component to component.

A component handling a voice signal is converted as a DSP program, and a component for calculating a control signal is programmed in the CPU control program. The voice component that receives the control signal also generates a part of the CPU control program.

Further, the CPU control program is divided into three programs of a tone generation control program, a key release control program and a regular control program in each task.

For the DSP program, each DSP 116
Although it is required for each, it is generated as one program in each DSP 116.

The work memory includes a general work memory for temporarily storing data during calculation and a wiring memory for input / output.

The process in the program generation unit 300 is DS
It is divided into a part that creates a P program and a part that creates a CPU control program, and the method of connecting templates and the method of allocating work memory are different.

(Regarding creation of DSP program) DS
Since only the voice component has the P program template, from the algorithm data,
First, create a list of audio components. When a plurality of the same components are used, a component ID, which is an identification number, is given.

Then, according to the list, the template of the DSP program and the symbol information are extracted from the component file. In the DSP program, the method of connecting the templates is arbitrary. That is, it may be in the order registered in the algorithm data or in the alphabetical order of the component names.

The work memory used by the DSP program includes a register memory, a coefficient memory and a delay memory.

All the above three memories are used as a general work memory, but the register memory is the only memory used for wiring.

Assignment of register memory for wiring is as follows.
First, set the output memory to 1 in the register memory area.
Assign to every other virtual address. Subsequently, the corresponding memory for input is assigned to the next virtual address for output. The register memory is a ring buffer, and in the DSP 116, the value obtained by adding the value of the virtual address and the value of the pointer that moves backward by one address in one cycle (sampling cycle) is stored in the register memory. It is an absolute address. In this way 1
Every cycle, the data will be sent to the next virtual address.

Therefore, in this way, when the audio signal passes from component to component, a delay of 1 sample is always generated. By doing so, it is possible to prevent the sample delay from occurring or not depending on the order in which the templates are connected.

In the case of a parallel component, which will be described later, all pins of the component connected to the component refer to the same register memory.

One word of the register memory (zero memory) from which "0" is always read is prepared in the register memory. When the input pin is not wired, this zero memory is assigned as the input memory.

If the output pin is not wired, a register memory other than the zero memory is assigned.
That is, if the output pin is unwired, no value will be read from this register memory, and no problem will occur as it is.

For example, D of the sound source algorithm shown in FIG.
The procedure for creating the SP program will be described with reference to FIG. 10. First, the templates of the components are arranged appropriately (in alphabetical order or in the order stored in the memory. In FIG. , The ones that use the register to be output are allocated to the virtual addresses one by one from the beginning, and then the ones to be input are allocated to the corresponding next virtual address.

That is, when Osc is stored (stored) in the virtual address of the register memory, Pan is allocated so as to load (load) the next virtual address. Hereinafter, similar virtual address allocation is performed.

Information for changing the left / right deviation of Pan and the pitch of Osc is read from the coefficient memory and calculated. Therefore, the values of various parameters can be controlled by writing the values from the CPU 104 to the coefficient memory. For this reason, the CPU 104
For example, it is necessary to give information indicating the address of the coefficient memory of Pan. That is, the CPU 104 directly specifies an address and writes the value when writing the value.

By performing the above processing, D
The SP program can be generated. Thus, D
When the generation of the SP program is completed, the completed DSP program is transferred to the DSP program memory. After the transfer is completed, the DSP immediately starts processing the audio signal by the program.

(Creation of CPU control program)
The CPU control program is generated from a voice component having a control component and control input pins and control output pins.

In the case of creating this CPU control program, the template and symbol information of the CPU control program are obtained from the related component file 900, as in the case of creating the DSP program.

The method of creating the program is almost the same as that of the DSP program, but is different in the following two points.

First, the first point is that the CPU control program consists of the following three programs.

(1) Periodic Control Program (Periodic Service Task) This is a program for calculating control signals at regular intervals.
Specifically, the time-varying parameter and the operator calculate a real-time variable parameter.

(2) Sound generation control program (assigner task) Performs control calculation for starting sound generation. Specifically, the process for giving parameters related to the sustain part from the attack part of the envelope to the envelope generator, and the LF
The initial phase of O is set.

(3) Key release control program (assigner task) The control signal for key release is calculated. Specifically, the parameters related to the release part of the envelope
Performs processing to give to the generator.

Next, the second point is that template connection is
It means to follow the order of signal flow as much as possible. This is based on the following reasons. That is, the CPU 104 executes the corresponding program according to the order in which the templates are linked, but the control program at the time of sounding or releasing the key is executed only once at the time of sounding / release of the key, so that the template is connected. If the order is reversed from the signal flow, the control signal calculated at the time of the previous sound generation / key release on the upstream side of the signal flow will be supplied to the downstream side at the new sound generation / key release, which is normal. This is because it does not operate normally.

Referring to FIG. 11, the method of connecting the templates in the order of signal flow as much as possible will be described. First, a list of components that will be the ends of the flow of control signals will be created. That is, it creates a list of components that have control input pins that are control signal input pins but no control output pins. Normally, an audio component such as Pan is terminated as shown in FIG. By tracing the flow of the control signal from the terminal component, the connection order of the templates is determined from the back.

When there are a plurality of termination components, the first termination component is focused on. The template for the component is the last in the control program. When there is only one terminal component (in the sound source algorithm shown in FIG. 9, Pan is the terminal component), the terminal component is focused on.

Then, the attention is paid to the component connected to the control input pin of the noticed terminal component, and the template is connected in front of the previous template (in the case of the sound source algorithm shown in FIG.
It becomes d. ).

By sequentially connecting the templates of the components connected to the input pins in this way, the templates can be connected according to the order of signal flow.

When a component has a plurality of control input pins, it traces upstream of one input pin until it reaches a component with no control input pin, returns to the last branched component, and continues to the next input pin. Just follow.

If the component of interest has a plurality of output pins, that is, if there is an output pin other than the one currently traced, there is a possibility that feedback is formed. Stop and set the flag corresponding to the component's current traced output pin.

In this way, when all the components that can be traced upstream from the termination component are traced, the same processing is repeated from the next termination component, and the above operation is repeated until there is no termination component.

For a component having a plurality of output pins, the process of tracing upstream is restarted when all the output pins are flagged.

If feedback is formed in the algorithm of the control signal, even if all the scans are finished, some control components in which the flags of the output pins are not set remain. Therefore, in the case of feedback, when the scan from the terminal component is completed, the control component whose processing has not been completed is searched and the upstream scan is performed.

By the above operation, the templates can be linked.

Next, a method of generating the above three programs from one algorithm will be described.

A component can have a plurality of control program templates. Which control program each template is linked to, and the template tense TplTens of the component file.
Specified by e. Its attributes are the following four types. (1) Connected as CYCL regular control program (2) Connected as PRON sounding control program (3) Connected as NOFF key release control program (4) NONE Connected to the same control program as downstream components

For example, lfoI (FIG. 43) is a component that can realize an LFO in which the initial phase at the start of sounding can be set, and has two templates. One is a template for generating a program to calculate the waveform in the regular service, which is specified in CYCL, and the other is a program for setting the initial phase when sounding (key pressing). Template to generate, PRO
Specified as N.

The program generation unit connects the program for calculating the waveform to the regular control program and the program for setting one initial phase to the sound generation control program according to the template attribute. As a result, the calculation processing of the waveform is periodically repeated, but the setting processing of the initial phase is performed only once at the time of sound generation.

Similarly, env (FIG. 49) and adsr (FIG. 50) are envelope envelope generating envelope signals.
Although it is a component that can realize a generator, it also sets a template for generating a program that sets the parameters related to the sustain part from the attack part of the envelope signal that is also generated when the sound is generated, and the parameters related to the release part when the key is released. A template for generating a program and a template for generating a program for calculating the waveform of the envelope signal in the regular service based on these set parameters are prepared.

The template attribute NONE is attached to the template of a general-purpose arithmetic program, and arithmetic system components, curve conversion system components, fader components, etc., which will be described later, have templates of this attribute. Since the processing of these components is performed in order to give control data to the control program of the component connected to the downstream, the template must be linked to the same control program as the component of the downstream. Specifically, the following processing is performed.

In the program generator 300, the system
A tense data buffer for storing tense data in the work area is prepared for each template. Temporal data represents the type of program that links the templates, and its possible values are CYCL, RRON, and NOF.
F, which corresponds to a regular control program, a sound generation control program, and a key release control program, respectively. Temporal data buffers are reserved for the number of control components used and are initially initialized with CYCL.

As described above, the program generator 300 sequentially connects the templates from the end component. In connecting the templates, first, the template tense TplTense of the template of interest is determined. Template tense TplTense is NON
If it is other than E, the template is connected to the program corresponding to the tense of the template. When the template tense TplTense of the focused component is NONE, the template is linked to the program corresponding to the tense indicated by the tense data buffer of the template.

In this case, when the focused component is the terminal, the tense data buffer is set to CYCL by the above initialization.

When the process related to the template connection is completed, the content of the tense data buffer of the upstream component is set in order to prepare the connection of the template of the upstream component connected to the control input pin of the focused component. . Specifically, if the pin tense PinTense of the focused component is other than NONE, the content of the tense of that pin is copied to the tense data buffer of the upstream component, and the pin tense PinT
If ense is NONE, the contents of the tense data buffer of the focused component are copied to the tense data buffer of the upstream component.

When the above processing is completed, the component of interest is moved to the upstream component, and the template connection processing is repeated.

For example, there are three control signal input pins for the component lfoI: a pin for inputting a rate for setting a frequency, a pin for inputting a waveform for setting a waveform, and a pin for inputting a phase for setting an initial phase. Of these, the rate for setting the frequency and the waveform for setting the waveform are referred to by a program for calculating the waveform (connected to the regular control program).

On the other hand, the phase for setting the initial phase is
It is referred to by the program that sets the initial phase when the key is pressed (connected to the sound generation control program).

Therefore, rate or wavefor
If the template attribute of the component connected upstream of m is NONE, it is linked to the periodic control program. On the other hand, if it is connected upstream of the phase, it is linked to the sound generation control program.

In this way, when tracing back from the downstream component to the upstream component through the pin connection, the data of the pin tense PinTense is used to inherit the link destination of the template.

In the end component that has no downstream component, since the tense data buffer is initialized by CYCL, the template of most of the components in which the pin tense PinTense and the template tense TplTense are NONE are periodically controlled. Be linked to the program.

(Regarding Processing of Trigger Pin) Next, the trigger pin (display device 60) arranged on the component
It is displayed in white on the screen of 0. ) Will be described.

The tone generation control program and the key release control program operate by being called by the assigner task. At this time, the tone generation channel number is passed as an argument. In the tone generation control program and the key release control program, processing for all tone generation channels is written, but only the portion corresponding to the tone generation channel number must be executed from that program.

Trigger pins are always arranged in the components related to the tone generation control program and the key release control program. On the contrary, only the component in which the trigger pin is arranged and the component connected upstream thereof are related to the tone generation control program and the key release control program.

Therefore, the program generation unit 300 checks the wiring of the trigger pin, and in the tone generation control program and the key release control program, issues a branch instruction for checking the tone generation channel number of the argument to the program corresponding to each channel. Put at the beginning of.

That is, the program is generated by connecting the templates corresponding to the components of interest and sequentially stacking them. If the component of interest has a trigger input pin at that time, the program generator 300 Remember the beginning positions of the pronunciation control program and the key release control program that have already been loaded.
Then, when all the control components connected to the component have been traced upstream, that is, when the templates have been stacked, a branch instruction is further connected to the head thereof. The branching condition is “when the sounding channel number of the argument does not match the sounding channel number of the assigner component upstream of the trigger input pin of the focused component”, and the branch destination is the remembered start position.

(Assignment of Work Memory) Next, assignment of work memory will be described with reference to FIG.

The work memory is assigned to the control program work area in the work RAM 110.
First, a general work memory and a work memory for output are allocated, and the dummy address in the concatenated template is replaced with the address of the allocated work memory.

Next, the work memory for input is assigned to the same address as the work memory for output wired, and the dummy address is replaced.

In the control program work area also, a zero memory is prepared as a work memory from which 0 is always read. If the input pin is not connected, this zero memory is allocated.

As described above, even if there is an unconnected pin, it is possible to generate a program even from an algorithm in the process of making it possible to perform verification and evaluation in the created range by not causing an error. ing.

(Regarding Writing of Coefficient from CPU 104 to DSP 116) Control of the audio component by the control signal is realized by writing the coefficient from the CPU 104 to the coefficient memory.

The template of the control program of the audio component having the control signal input includes an instruction to write data from the CPU 104 to the coefficient memory of the DSP 116. In the template, the address of this coefficient memory is also written as a dummy.

Since the coefficient memory is assigned when the DSP program is generated, the assigned address replaces the dummy address of the template.

The control program can be generated by performing the above processing.

When the generation of the CPU control program is completed in this way, the completed CPU control program is transferred to the control program area of the work RAM 110.
Immediately after the transfer, the CPU 104 starts the control processing of the DSP 116 by the control program. This allows the operator to realize an electronic musical instrument having a desired algorithm.

[Component File 900] Next, a typical component among the components stored in the component file 900 is shown in FIG.
2 to 61, description will be given below.

Although alphanumeric characters are used for component names, voice component names are capitalized and control component names are capitalized.

[Voice Component] Filter System (1) TVF (FIG. 12) This component realizes the function of the time-varying filter.
Time-varying filters are often used in synthesizers, but only the cut-off frequency is time-varying, and the resonance (Q) can be fixed or controlled manually. In the system according to the present invention, two parameters of cutoff frequency and resonance can be freely time-varying. That is, LFO and envelope can be applied. You can also change the resonance according to note-on velocity, note-off velocity, aftertouch, etc.

The cut-off frequency is given as data in cent units from the outside so that the control can be performed easily.
It is converted to an actual cut-off frequency inside the TVF and used for control. This also applies to the center frequency of TVEQ described next.

(2) TVEQ (time-varying equalizer) (Fig. 1
3) It is a component that realizes the function of the time-varying equalizer. The center frequency, Q, and gain can be controlled.

Gain system (1) TVA (FIG. 14) This is a component that realizes the function of the time-varying amplifier, and the level can be controlled.

(2) Mix (FIG. 15) CMix (FIG. 16) This is a component that realizes the function of the mixer. Mi
x mixes at a constant volume. The CMix can independently control the level of each input according to the corresponding control signal.

(3) Pan (FIG. 17) This is a component that realizes the function of the pan pot.
L, R according to the supplied localization information (panpot)
The level ratio of the two outputs is controlled.

Oscillator (1) Osc (FIG. 18) This component realizes the function of the oscillator and generates a triangular wave. Pitch and level (le
vel) can be controlled. It should be noted that the pitch is given from the outside as data in units of cents with a predetermined pitch as a reference so as to facilitate control, and is converted into an actual frequency inside the Osc for use in control. This is the Pcm described below.
The same applies to WG.

The triangular wave can be generated by a simple circuit as shown in FIG. Since the value of the audio signal is represented by 2's complement, a sawtooth wave (FIG. 19B) is generated by accumulating pitches for each sample. The sawtooth wave is passed through the absolute value circuit (ABS) (Fig. 19).
(C)), a shift results in a triangular wave (FIG. 19 (d)).

From the CPU, pitch and level can be set. pitch is the frequency of the triangular wave,
l controls the amplitude of the triangular wave.

(2) PcmWG (FIG. 20) This is a component that realizes the function of the oscillator for reading the PCM waveform. This is PCM, not DSP116
The corresponding channel of the waveform reading device 114 is controlled. Like Osc, the pitch and level can be controlled.
In addition, the reading of the PCM waveform is started from the head position of the waveform to be read in response to the trigger input (trig).

In the tone generation control program, the pitch data corresponding to the input control signal is given to the corresponding channel of the PCM waveform reading device 114, and the generation start of the waveform is instructed.

In the regular control program, the pitch data corresponding to the input control signal value is continuously controlled.

Delay system (1) Del (FIG. 21) DelSI (FIG. 22) DelF (FIG. 23) This component realizes the delay function, and the delay time can be set. DelSI is a delay
It is a delay component that interpolates when the time is changed. Even if the input delay time changes, interpolation is applied and the actual delay time changes gradually,
The generation of noise due to changes in the delay time is prevented. The time constant (tc) of interpolation can also be set. Also,
DelF has feedback and can control the feedback level (fbk).

(2) Chorus (FIG. 24) This is a component that realizes the chorus function. The delay time of the delay circuit for realizing the chorus effect is controlled by the delay time (delay). The rate and the depth control the period and the size of the LFO that modulates the delay time, and the undulation period and the depth of the chorus effect are controlled.

(3) Reverse (FIG. 25) This is a component for realizing the reverb function. The level of the original audio signal to which the reverb effect has not been added is controlled by dry, and the effect (ef
x) controls the level of the audio signal to which the reverb effect is added.

Others (1) Output (FIG. 26) This is a component that realizes the function of outputting an audio signal to the D / A converter 124. The D / A output selector 118 is set so that the D / A output of the DSP 116 in which this component is placed is sent to the D / A converter 124. This setting is done at compile time. When using a plurality of DSPs 116, they can be placed in only one of the windows.

(2) PPara (FIG. 27) This is a parallel component and realizes a function of distributing an audio signal to a plurality of components.

(3) Sel (FIG. 28) An audio signal selector component, which realizes the function of selecting one from a plurality of audio signals.

(4) Ring (FIG. 29) This is a component that realizes the function of the ring modulator.

[Control Component] Input System (1) fader (FIG. 30) This is a fader component for the operator to manually set the value of the control signal. It has the shape of a fader (slide volume), and you move the knob with the mouse to set the desired value. The value can also be increased or decreased by clicking the increment button or decrement button with the mouse.

An arbitrary character string can be displayed on the upper part and can be used as a parameter name. The control program of the fader component simply fetches the value from the work memory holding the value and passes it to the next component. When the fader is operated, the algorithm creating section 200 sends the changed value to the sound source section 100. CP
The value is written to the work area via U104.

(2) cc (FIG. 31) pBend (FIG. 32) chAft (FIG. 33) cc is a continuous control message,
pBend is a pitch bend control message and chAft is an after touch control message, and the MIDI receiving task outputs the data written in the buffer as a control signal.

The control signal input pin ctrl # of cc is for inputting the control number of the continuous control message of the MIDI information output from this component. For example, if you connect a fader component to this pin and enter "40 (decimal)" as the value, hold 1 corresponding to number 40 from this component in the received MIDI information.
(Damper) information will be output.

(3) asgnNm (FIG. 34) asgnNL (FIG. 35) asgnN (FIG. 36) This is an assigner component which converts MIDI information of a note assigned to a sounding channel into a control signal. The sounding channel number is entered in m. Pitch
#: Converts the note number to pitch data in cents based on a specified pitch and outputs it), and outputs on velocity, off velocity, and polyphonic after touch as control signals. Also, it has a trigger output pin that outputs a trigger signal at the time of sound generation (at the time of note on) and at the time of releasing the key (at the time of note off). The asgnNL corresponds to the sounding channel assigned to the last-arriving note.

AsgnN is a virtual component for creating an algorithm of asgnNm that can be used only in the polyphonic window. When compiled, it operates as an independent asgnNm component for each channel.

The above-mentioned assigner component includes
It has two roles. One is to set a correspondence relationship between a component that operates in response to key press / release such as an envelope generator and a channel assigned by the assigner task. When creating an algorithm, the operator connects the trigger pin of the assigner component and the control component that operates in response to key press / release such as the envelope generator, so that the channel assigned by the assigner and the control component concerned can be connected. Instruct the correspondence with.

The program generating section generates a program so that the control program of the control component corresponding to the channel where the key depression or key release occurs is executed according to the processing of the trigger pin described above.

Another role is note number, on velocity, off velocity, polyphonic
The role is to take out MIDI information as a control signal, which independently responds to key depression or key release such as after touch.

The note number and on-velocity are stored in the corresponding channel buffer in the key press processing of the assigner task, and the off-velocity in the key release processing of the assigner task.

The control program of the assigner component reads these data from the channel buffer and outputs them as control signals.

A polyphonic after touch buffer is provided for each note number,
When the polyphonic after touch information is received, it is stored in the position of the corresponding note number of the polyphonic after touch buffer in the MIDI reception task. The control program of the assigner component reads the note number assigned to that sounding channel from the channel buffer, and starts the polyphonic after touch data from the position corresponding to the note number in the polyphonic after touch buffer. Is read out and output as a control signal.

Operation system (1) add (FIG. 37) sub (FIG. 38) mul (FIG. 39) add is a component that realizes the function of an adder that obtains the sum of two control signals. Sub is a component that realizes the function of a subtractor that obtains the difference between two control signals. mul is a component that realizes the function of a multiplier that obtains the product of two control signals.

(2) max (FIG. 40) min (FIG. 41) max is a Maxim component that compares the magnitudes of two control signals and obtains the larger one. min is a minimum value that compares the magnitude of two control signals and obtains the smaller one.
It is a component.

Oscillator system (1) lfo (Fig. 42) lfoI (Fig. 43) lfoPG (Fig. 44) lfoWG (Fig. 45) lfoPS (Fig. 46) These realize the function of the LFO that generates a periodic waveform for modulation. This is an LFO component that is used for vibrato, tremolo, etc. and can set the cycle and waveform. lf
The oI has a trigger input and can reset the phase when the key is pressed to the phase corresponding to the value of the control signal provided to the initial phase pin.

LfoPG and lfoWG are normal lf
The function of o is divided into two. lfoPG generates a phase signal that changes with time at a speed corresponding to the rate, and lfoWG generates a waveform from the phase. lfo
The PS can be arbitrarily shifted in phase. As shown in FIG. 47, these are used to perform various calculations and processes at the time of the phase signal, so that a more multifunctional lfo component can be realized. For example, FIG. 48 shows an algorithm of two LFOs whose phase difference can be set freely. In this case, the control signals generated by the two LFOs have the same waveform in the same cycle but have a phase difference that can be set arbitrarily.

(2) env (FIG. 49) adsr (FIG. 50) These are envelopes that generate an envelope signal.
It is a component that realizes the function of the generator.
The env is a method of setting the level (L) and time (T) of five break points.

It should be noted that L3 is the sustain level,
The calculation of the parameters from L0 to L3 is entered into the tone generation control program. T4 (release time) and L4 (release end level) enter the key release control program. adsr is a method often used in analog synthesizers to set attack time, decay time, sustain level, and release time.

(3) rnd (FIG. 51) This is a random component that realizes the function of generating random numbers, and is used as noise.

(4) noise1_f (FIG. 52) This is a noise component for realizing the function of generating noise having a frequency characteristic of 1 / f.

Curve conversion (1) 12e (FIG. 53) This is a linear / exponential conversion component that realizes the function of performing linear / exponential conversion. This component converts the characteristic curve of the control signal value into an exponential curve.For example, when setting the delay time of the delay component with a fader, insert this component between the fader and the delay time pin. By inserting, it becomes possible to easily perform the setting when the delay time is small and the setting when the delay time is large.

(2) c2stp (FIG. 54) This is a component for converting the pitch control signal supplied from the assigner component into a delay time (delay step number) corresponding to the pitch cycle. As shown in FIG. 55, when a delay component is combined with a mixer component, a parallel component, etc. to form a feedback path to form a comb filter, this component is used to adjust the pitch control signal to the corresponding delay time. It is possible to correspond to the musical sound that generates the resonance frequency of the comb filter by converting to the.

(3) limit (FIG. 56) This component realizes the function of holding the value of the control signal within a certain range.

Others (1) constN N is "1" (Fig. 57), "0.5" (Fig. 58), "-".
It is a component that realizes the function of taking a value such as "1" (FIG. 59) or "-0.5" (FIG. 60) and giving a constant to the control signal.

(2) cPara (FIG. 61) This is a parallel component that realizes the function of distributing a control signal to a plurality of components.

[0276]

Since the present invention is constructed as described above, it has the following effects.

Storage means for storing component information for constructing a plurality of components relating to tone generation for each component corresponding to the component, selection means for selecting the component information stored in the storage means, and selection Input means for inputting connection state information indicating an arbitrary connection state between components corresponding to the component information selected by the means, and constructing the component and connection state based on the component information and connection state information, and input performance Programmable processing means for processing information and generating musical tones,
A component corresponding to the component information selected by the selecting means in the processing means by automatically programming the processing means based on the component information selected by the selecting means and the connection state information input by the input means. And the programming means for constructing the connection state indicated by the connection state information input by the input means, the component information for constructing the component stored in the storage means by the selection means and the input means is provided. Along with the selection, it is possible to input connection state information indicating the connection state between the components corresponding to the selected component information.

Then, the programming means controls the processing means to connect the components corresponding to the component information in the connection relation based on the connection status information and automatically generate the DSP and CPU programs based on the connection status information. Then, an electronic musical instrument having a desired sound source algorithm is realized.

Therefore, the simple operability in creating the sound source algorithm of the conventional modular synthesizer can be realized by the digital sound source, and the operator of the electronic musical instrument himself can
Arbitrary sound source algorithms for electronic musical instruments can be easily constructed, so that it is possible to realize an electronic musical instrument that is stable and stable in space and cost, and has excellent operability. You can

Further, when the electronic musical instrument according to the present invention is used,
It will also be possible to control musical sounds like never before. For example, parameters such as chorus and reverb that were previously set manually (chorus depth, delay time,
(Reverb time, etc.) using the output signal of the envelope generator, note-on velocity, note-off velocity, after touch, etc.
It is also possible to control so as to change automatically.

[Brief description of drawings]

FIG. 1 is an overall system configuration diagram of an electronic musical instrument according to an embodiment of the present invention.

FIG. 2 is a block configuration diagram of a sound source unit.

FIG. 3 is an explanatory diagram of a display screen of a display device.

FIG. 4 is an explanatory diagram schematically showing the data structure of a component file.

FIG. 5 is an explanatory diagram showing a configuration of components,
(A) is explanatory drawing which shows the method of determining the magnitude | size of a component, (b) shows a control signal pin and a sound signal pin, (c) shows a trigger signal pin, (d) is a sound signal pin. 6 shows a pattern of a component (audio component) having only a signal, (e) shows a pattern of a component having only a control signal pin and a trigger signal pin (control component), and (f) shows a pattern of the control signal pin and a trigger signal 3 shows a pattern of components with audio signal pins.

FIG. 6 is an explanatory diagram of a display screen of a command list window.

FIG. 7 is an explanatory diagram of a screen showing a process when a plurality of DSPs are used.

FIG. 8 is an explanatory diagram of a screen showing polyphonic processing.

FIG. 9 is an explanatory diagram of a connection state of components showing an example of a sound source algorithm.

FIG. 10 is an explanatory diagram of a DSP program generation process.

FIG. 11 is an explanatory diagram of a CPU control program generation process.

FIG. 12 shows a display state of the TVF on the screen of the display device.

FIG. 13 shows a display state of TVEQ on the screen of the display device.

FIG. 14 shows a display state of TVA on the screen of the display device.

FIG. 15 shows a display state of Mix on the screen of the display device.

FIG. 16 shows a display state of CMix on the screen of the display device.

FIG. 17 shows a display state of Pan on the screen of the display device.

FIG. 18 shows a display state of Osc on the screen of the display device.

FIG. 19 (a) shows a circuit configuration for generating a triangular wave, and FIG. 19 (b) shows a waveform of a generated sawtooth wave.
(C) shows the waveform after passing the sawtooth wave through the absolute value circuit, and (d) shows the waveform of the triangular wave obtained by shifting the waveform of (c).

FIG. 20 shows a display state of PcmWG on the screen of the display device.

FIG. 21 shows a display state of Del on the screen of the display device.

FIG. 22 shows a display state of DelSI on the screen of the display device.

FIG. 23 shows a display state of DelF on the screen of the display device.

FIG. 24 shows a display state of Chorus on the screen of the display device.

FIG. 25 shows a display state of the Reverse on the screen of the display device.

FIG. 26 shows a display state of Output on the screen of the display device.

FIG. 27 shows a display state of PPara on the screen of the display device.

FIG. 28 shows a display state of Sel on the screen of the display device.

FIG. 29 shows a display state of Ring on the screen of the display device.

FIG. 30 shows a display state of a fader on the screen of the display device.

FIG. 31 shows a display state of cc on the screen of the display device.

FIG. 32 shows a display state of pBend on the screen of the display device.

FIG. 33 shows a display state of chAft on the screen of the display device.

FIG. 34 shows a display state of asgnNm on the screen of the display device.

FIG. 35 shows a display state of asgnNL on the screen of the display device.

FIG. 36 shows a display state of asgnN on the screen of the display device.

FIG. 37 shows a display state of add on the screen of the display device.

FIG. 38 shows a display state of sub on the screen of the display device.

FIG. 39 shows a display state of mul on the screen of the display device.

FIG. 40 shows a display state of max on the screen of the display device.

FIG. 41 shows a display state of min on the screen of the display device.

FIG. 42 shows a display state of lfo on the screen of the display device.

FIG. 43 shows a display state of lfoI on the screen of the display device.

FIG. 44 shows a display state of lfoPG on the screen of the display device.

FIG. 45 shows a display state of lfoWG on the screen of the display device.

FIG. 46 shows a display state of lfoPS on the screen of the display device.

[FIG. 47] lfoPG, lf on the screen of the display device
The display state when combining oPS and lfoWG is shown.

[Fig. 48] Fig. 48 is an explanatory diagram of a component connection state showing an algorithm for realizing the functions of two LFOs in which the phase difference can be freely set.

FIG. 49 shows a display state of env on the screen of the display device.

FIG. 50 shows a display state of adsr on the screen of the display device.

FIG. 51 shows a display state of rnd on the screen of the display device.

[Fig. 52] noise1_f on the screen of the display device
Shows the display state of.

FIG. 53 shows a display state of 12e on the screen of the display device.

FIG. 54 shows a display state of c2stp on the screen of the display device.

[Fig. 55] Fig. 55 is an explanatory diagram of a connection state of each component illustrating an example of an algorithm using c2stp.

FIG. 56 shows a display state of the limit on the screen of the display device.

FIG. 57 shows a display state of const1 on the screen of the display device.

FIG. 58 shows a display state of const05 on the screen of the display device.

FIG. 59 shows a display state of const-1 on the screen of the display device.

FIG. 60: const-05 on the screen of the display device
Shows the display state of.

FIG. 61 shows a display state of cPara on the screen of the display device.

[Explanation of symbols]

 100 Sound Source Section 200 Algorithm Creation Section 300 Program Generation Section 400 MIDI Keyboard 500 Sound System 600 Display Device 700 Keyboard 800 Mouse 900 Component File 1000 PCM Waveform Data File

Front page continued (72) Inventor Takashi Toyomura 1-4-16 Dojimahama, Kita-ku, Osaka, Osaka Prefecture Roland Co., Ltd. (72) Inventor Shinobu Katayama 1-4-16 Dojimahama, Kita-ku, Osaka, Osaka Roland Stock In-house (72) Inventor Tadashi Yamanaka 1-4-16 Dojimahama, Kita-ku, Osaka City, Osaka Prefecture Roland Co., Ltd. (72) In-house Toshiro Yamabata 1-4-16 Dojimahama, Kita-ku, Osaka City, Osaka Prefecture Roland Co., Ltd. (72) Inventor Adrian Bruce 1-4-16 Dojimahama, Kita-ku, Osaka-shi, Osaka Prefecture Roland Corporation

Claims (1)

[Claims] 1. A storage unit for storing component information for constructing a plurality of components relating to tone generation for each of the components in association with the component, and selecting the component information stored in the storage unit. Selecting means, input means for inputting connection state information indicating an arbitrary connection state between the components corresponding to the component information selected by the selecting means, the component information and the connection state information based on the connection state information Build the component and the connection state,
Programmable processing means for processing input performance information to generate musical tones, and to the processing means based on the component information selected by the selecting means and the connection state information input by the input means. Programming that causes the processing means to build the connection state indicated by the component corresponding to the component information selected by the selection means and the connection state information input by the input means by automatically programming An electronic musical instrument having means.