JP3152198B2 - Music sound generation method and music sound generation device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a musical tone generating method and apparatus having a waveform memory sound source for generating musical tones by reading waveform sample data stored in a memory.

[0002]

2. Description of the Related Art In a conventional tone generating apparatus having a waveform memory sound source, when performance information is input, for example, a sound source driver processing program is read from a program memory and executed by an arithmetic processing means (CPU). .
As a result, the sound source parameter information is transmitted, and this parameter information is stored in the sound source register. In the waveform memory tone generator, tone generation processing for generating tone data based on the parameter information stored in the tone generator register is executed. That is, the waveform sample data is read from the waveform memory in which the waveform sample data is stored, and an envelope is added to the read waveform sample data, and an effect is further applied to generate musical sound data.

[0003]

Here, a program such as a sound source driver processing program executed by the CPU and waveform sample data read by the waveform memory sound source are stored in a read only memory (ROM). There are cases. This is because if the ROMs can be integrated, the circuit can be reduced in size as a whole, and it is advantageous in terms of cost. Then RO
The access time of M is generally RAM (Random Access
Memory) is longer than the access time.
However, there is a problem in that if a program is read and executed from a ROM, the performance of the CPU is reduced. Further, the CPU and the waveform memory tone generator share a ROM on which a program and waveform sample data are stored on a memory bus. For this reason, when a memory access conflict occurs between the waveform memory sound source unit and the CPU, for example, a wait is inserted into the CPU to
There is a problem that the performance of U is reduced. As described above, the fact that the CPU is weighted and the response is slow is a serious problem particularly for a musical sound generator, an electronic musical instrument, and the like, which require real-time performance for the performance operation and accuracy for the progress of the music.

Accordingly, the present invention provides a tone generating method and a tone generating method in which a program is read from a low-speed ROM and stored in a high-speed memory when the capacity of the high-speed memory has room, thereby preventing a decrease in CPU performance. It is intended to provide a device.

[0005]

In order to achieve the above object, a musical sound generating method according to the present invention comprises: a plurality of programs read and executed by an arithmetic processing unit in a read-only memory shared on a memory bus; The waveform memory sound source unit stores waveform sample data that is read out to generate a musical tone waveform. The waveform memory sound source unit stores the read-only memory.
Read waveform sample data from memory to generate musical tone waveform
The arithmetic processing unit detects a capacity of a high-speed memory in which a work area area used when executing the program is set, and detects a number of the programs according to the detected high-speed memory capacity. When the program is copied from the read-only memory to the high-speed memory and the program is executed , the waveform processing sound source unit reads the program copied by the arithmetic processing unit into the high-speed memory.
Used to read waveform sample data from the
The data is read out and executed by using a bus different from the bus used .

A musical sound generating apparatus according to the present invention, which can achieve the above object, comprises an arithmetic processing unit, a waveform memory tone generator for generating musical sound data, a plurality of programs executable by the arithmetic processing unit, and a read only memory that is shared on the memory bus by the said processing unit, wherein the waveform memory tone generator section during tone waveform generation waveform sample data read out is stored the waveform memory tone generator section, before
The waveform memory sound source section receives the waveform from the read-only memory.
The waveform sample data is read out to generate a musical tone waveform, the capacity of a high-speed memory in which a work area for the arithmetic processing unit is set at startup is detected, and the number of the programs corresponding to the detected high-speed memory capacity is determined. From the read-only memory to the high-speed memory,
When executing the program, the arithmetic processing unit copies the program copied to the high-speed memory into the waveform.
The memory sound source section reads the waveform sampled from the read-only memory.
Use a bus different from the bus used to read
It is used to read and execute.

According to the present invention, a program is copied from a low-speed memory to a high-speed memory when the capacity of the high-speed memory is sufficient. And can be executed by reading from the memory. Therefore, the program can be executed at high speed, and the performance of the CPU can be improved. At this time, the CPU accesses the high-speed memory, and the sound source section stores the R in which the waveform sample data is stored.
Since the OM is accessed, contention of memory access between the CPU and the sound source section does not occur. Therefore, it is possible to prevent the performance of the CPU from deteriorating, and the CPU can perform arithmetic processing at a higher speed.

[0008]

FIG. 1 is a block diagram showing a configuration example of an embodiment of a musical sound generating apparatus according to the present invention in which a musical sound generating method according to the present invention is executed. In FIG. 1, reference numeral 1 denotes a sound source LSI in which an arithmetic processing unit (Central Processing Unit: CPU) 11 and a waveform memory sound source unit and its peripheral circuits are formed on one chip, and 2 denotes a waveform accessed by the sound source LSI 1 This is a WAVE & PROGRAM ROM in which sample data and programs are stored. Reference numeral 3 denotes a DELAY & WORK RAM which is used as a delay unit when performing an effect process for giving an effect to a musical tone in the waveform memory sound source unit and in which a work area used by the CPU 11 is set. It is preferable to use a RAM (DRAM). Further, 4 is an analog waveform input means such as a microphone, and 5 is an ADC (Analog Digital C) which converts an analog signal input from the analog waveform input means into a digital signal and inputs the digital signal to the sound source LSI 1
onverter).

Further, reference numeral 6 denotes D for converting digital tone data output from the tone generator LSI 1 into analog tone signals.
AC (Digital Analog Converter), 7 is DAC
6 is a sound system that amplifies and emits the analog musical tone signal converted by the analog tone signal 6. 8 is an external MIDI
MIDI terminal for inputting / outputting MIDI signals from devices, 9
Is a keyboard provided with a plurality of keys on which a user performs a performance operation;
Is a panel display for displaying various information connected to the parallel input / output port 15, a panel operator for giving various instructions of the tone generator by a user's operation, or an external CPU that can replace the CPU 11. is there.

Further, reference numeral 11 denotes a CPU for controlling the whole operation of the tone generator, and 12 denotes an elapsed time during the operation of the tone generator or an automatic performance or an automatic accompaniment which generates a timer interrupt at a specific interval. A timer 13 for performing time management, envelope control, effect control, and the like, for taking in a MIDI message input from an external MIDI device via the MIDI terminal 8 into the tone generator LSI 1 and for converting a MIDI message generated in the tone generator LSI 1 into MIDI. A MIDI interface 14 for outputting to an external MIDI device via a terminal 8 is a serial input / output (serial I / O) 14 for receiving key-on information of note-on and note-off of the keyboard 9 and after-touch data as serial data. Port, 15
Is a parallel input / output (parallel I / O) for outputting data for controlling a panel display, receiving operation data of a panel operator, and exchanging data with an external CPU.
Port. Reference numeral 16 denotes a WAVE & PROGRAM ROM 2 (hereinafter referred to as WP-R) from the CPU 11 and the read circuit 21.
The access management units A and 17 that manage access to the OM2) include a DELAY & WORK RAM3 (hereinafter, DW-RA) from the CPU 11 and the DSP with mixer 23.
M3), and the access management unit A16 and the access management unit B17 operate as selectors that give priority to the access from the readout circuit 21 and the DSP with mixer 23. Further, reference numeral 18 denotes a bus line for connecting each unit.

Reference numeral 20 denotes tone control data for controlling the tone generated in each tone generation channel in the waveform memory tone generator.
A sound source register for storing mixer control data for controlling the DSP with mixer 23 and DSP control data,
1 generates read addresses for each of a plurality of tone generation channels (for example, 32 channels) based on the frequency number parameter and tone color parameter stored in the tone generator register 20, and generates a number necessary for interpolation processing based on the read addresses. The readout circuit reads out the waveform sample data of each tone generation channel from the WP-ROM 2 and performs an interpolation process on the read out waveform sample data of each tone generation channel. Reference numeral 22 denotes an EG assigning unit that assigns a volume envelope or the like based on the envelope parameter of the channel in the tone generator register to the waveform sample data of each sounding channel output from the readout circuit 21, and 23 denotes waveform sample data of each sounding channel and The waveform data from the analog waveform input unit 4 is mixed in a plurality of ways based on the mixer control data in the tone generator register 20, and a part of the mixing output is converted into the DSP in the tone generator register 20.
A DSP with a mixer (Digita) that adds effect processing such as chorus, reverb, variation, etc. based on the control data and outputs it as tone data for each sampling period.
l Signal Processor). The waveform memory sound source unit includes a sound source register 20, a readout circuit 21, an EG adding unit 2
2, composed of a DSP 23 with a mixer.

The operation of the tone generator according to the present invention thus constructed will be described. Key press information is fetched from the keyboard 9 via the serial I / O port 14, MIDI
When the sounding timing of the event comes, the CPU 11 executes the sound source driver process and sends out the sound source parameters to the sound source register 20. At this time, if a note-on event is input, the CPU 11 first assigns channels, generates note numbers, timbre information (waveform addresses), and velocities (envelopes) so that a new musical tone can be generated based on the note-on events. Parameters), modulation parameters, etc., are set in the channel positions assigned by the sound source register 20. When a note-off event is input, an envelope parameter or the like that can stop sound generation based on the note-off event is set in the tone generator register 20. Note that these instructions are performed by the CPU 11 executing a sound source driver processing program stored in the WP-ROM 2.

In the waveform memory tone generator, a tone generation process is executed at predetermined timings. In this tone generation processing, the address readout circuit 21 is based on the accumulated value of the frequency information and the waveform address corresponding to the note number of each channel set in the tone generator register 20 in accordance with the time division channel timing obtained by equally dividing one sampling period. WP
Create a read address for ROM2. Then, by accessing the WP-ROM 2 at a timing different from the time-division channel timing according to the integer part of the created read address, the waveform sample data required for each channel is read every sampling cycle and stored in the buffer. Is stored. In this case, the required number of waveform sample data for calculating by interpolation of the waveform sample data according to the decimal part of the created read address is prepared in the buffer. Then, from the waveform sample data of each channel prepared in the buffer, the interpolation sampled waveform sample data is calculated according to the decimal part of the created read address. The interpolation process is performed for each channel at the time-division channel timing.

Here, the buffer is provided independently corresponding to each channel.
Since data read in a certain sampling period can be used in the next sampling period, the WP-ROM 2 is accessed by the same number of times as the number of read addresses advanced in each sampling period, and the number of waveform samples is increased. Data is read and supplied to the buffer. Therefore, the number of accesses for each sampling cycle is increased in a channel having a large frequency information, and the number of accesses is decreased in each sampling cycle for a small channel.

In order to prepare the necessary number of waveform sample data in the buffer in accordance with the decimal part of the read address in order to calculate the waveform sample data by interpolation, the waveform newly read from the WP-ROM 2 in each channel is prepared. The number of sample data, that is, the read circuit 21
In, the number of accesses to the WP-ROM 2 in one sampling cycle is calculated for each channel. The number of accesses required for each channel is accumulated for all channels, and WP
-The total number of accesses to the ROM 2 is calculated. When the total number of accesses is large, the priority of the waveform reading process is high, so that the CPU 11
In some cases, a wait cycle is inserted when accessing the P-ROM 2, and the performance of the CPU 11 may be reduced. Therefore, to prevent this,
WP-R by reading circuit 21 in sampling period
If an upper limit is set for the total number of accesses to OM2 and the calculated total number of accesses exceeds the upper limit,
The read circuit 21 reduces the total number of accesses by changing the mode of the interpolation process from interpolation to extrapolation.
Executed in

The waveform sample data of each channel after the interpolation processing output from the readout circuit 21 is provided with an envelope corresponding to the volume information and timbre information of the channel set in the sound source register 20 by the EG providing unit 22. You. Further, audio data and the like digitized by the ADC 5 input from the analog waveform input unit 4 are mixed with the waveform sample data of each channel to which the envelope is added in the DSP with mixer 23. At this time, mixing is performed in a plurality of ways based on the mixer control data in the tone generator register 20, and a part of the mixing output is subjected to effect processing such as chorus, reverb, and variation based on the DSP control data in the tone generator register. As a result, tone data is generated for each sampling period. DSP2 with this mixer
In 3, when the tone data needs to be delayed during the effect processing, the tone data is written to the DW-RAM 3 and read out after a predetermined time, so that the DW-RAM 3 is used as delay means. The tone data output from the DSP with mixer 23 at every sampling cycle is DA
C6, the analog tone signal converted by the DAC 6 is converted into an analog tone signal.
Is amplified and pronounced.

When a MIDI message is input from an external MIDI device via the MIDI interface 13, a tone generator driver process is executed in response to the MIDI message, whereby tone generator parameters are set in the tone generator register 20, and the above-mentioned keyboard is set. Similarly to the tone generation based on the performance information from the tone generator 9, the tone based on the MIDI message is generated by the waveform memory tone generator and is emitted from the sound system 7. Further, the parallel I / O port 15 outputs image data to be displayed on the panel display, and outputs a scan signal for scanning a panel switch and capturing its operation signal. Note that the parallel I / O port 15 can be connected to an external CPU. In this case, the operation of the internal CPU 11 is stopped, and each block of the sound source LSI 1
It is controlled by an external CPU through / O15.

By the way, the programs stored in the WP-ROM 2 include a main routine program (MAIN) for calling a program corresponding to a starting factor and executing the processing, and a main routine program (MAIN) for calling when a performance event occurs. A sound source driver processing program (TG_T) which is a routine for supplying sound source parameters corresponding to the performance event such as MIDI to the sound source register 20.
DRV), which is called from the main routine program when an operation event for controlling automatic performance or a timer interrupt event occurs, and is a routine for changing the setting of automatic performance or generating a performance event corresponding to the time of music data. A song processing program (SG_PL
Y), a routine that is called from the main routine program when an operation event for controlling automatic accompaniment or a timer interrupt event occurs, and that changes the setting of automatic accompaniment or generates a performance event corresponding to the time of the accompaniment pattern data. Style processing program (SL_
PLY), there is another processing program (OTHER) which is called from the main routine program when a panel switch operation event occurs and is a routine for editing timbre data and setting processing of the entire apparatus. Among them, the heaviest processing is the sound source driver processing, and the sound source driver processing program is stored in the WP-ROM 2 having a slow access time.
When reading and executing the program, much time is spent executing the sound source driver processing program, and the performance of the CPU 11 is degraded.

Therefore, the access time D
If there is free space in the W-RAM 3, the WP-ROM 2
If the sound source driver processing program is preferentially read from the DW-RAM 3 and copied to the DW-RAM 3, the DW-RA
It can be read from M3 and executed. in this case,
The sound source driver processing program can be executed at high speed.
The performance of U11 can be improved. Furthermore, if not only the sound source driver processing program but also other programs are read from the WP-ROM 2 and copied to the DW-RAM 3 according to the free space of the DW-RAM 3, the performance of the CPU 11 can be further improved. become able to. In addition, if the main routine program does not operate at high speed first, no matter how fast other routines can operate, the overall operation cannot be accelerated. Therefore, the main routine program has the highest priority.

The tone generator of the present invention has a DW
A start routine for reading a main routine program, a sound source driver processing program, and other programs from the WP-ROM 2 and copying the read program to the DW-RAM 3 according to the free space of the RAM 3; By executing at the time of raising, the CPU 11
Try to improve performance. Note that W
Various programs stored in the P-ROM 2 are relocatable programs and can be executed no matter where they are stored. So D
An example of a mode in which a main routine program, a sound source driver processing program, and other programs are read from the WP-ROM 2 and copied to the DW-RAM 3 according to the free space in the W-RAM 3 will be described with reference to a memory map shown in FIG. This will be described below.

FIG. 2A shows a memory map of the WP-ROM 2, an area for storing waveform sample data (WAVE) for a plurality of timbres, a main routine program (MAIN), and a sound source driver processing program (TG).
_DRV), a song processing program (SG_PLY),
An area for storing a program including a style processing program (SL_PLY) and other processing programs (OTHER) is set. FIG. 2B shows a DW-
The DRAM constituting the RAM 3 is not added and the DW-R
A state where the memory capacity of OM3 is very small is shown. In this case, DELA used as a delay means is shown.
Y area and a small work area (WOR
K) An area is set. Then, FIGS. 2C and 2D
(E) and (f) show the memory map of the DW-RAM 3, showing a mode in which the DRAM capacity of the DW-RAM 3 is gradually increased and the memory capacity is increased.

Here, the D constituting the DW-RAM 3
If the RAM is increased and the memory capacity is reduced to the small memory capacity shown in FIG.
An AY area, a slightly enlarged work area (WORK) area of the CPU 11, and an area for storing a copied top-priority main routine program (MAIN) are set. That is, when the memory capacity is as shown in FIG. 2C, only the main routine program (MAIN) is copied to the DW-RAM 3. Furthermore, DW-RAM
3 is further added to the DRAM shown in FIG.
In the case where the memory capacity is increased to the capacity indicated in (1), a DELAY area used as delay means, a larger work area (WORK) area of the CPU 11, a copied main routine program (MAIN) and a sound source driver processing program ( An area for storing (TG_DRV) is set. That is, when the memory capacity shown in FIG. 2D is set, the main routine program (MAIN) and the sound source driver processing program (TG_DRV) are DW-R
Copied to AM3.

Further, when the DRAM constituting the DW-RAM 3 is further expanded to have the memory capacity shown in FIG.
An ELAY area, a larger work area (WORK) area of the CPU 11, a copied main routine program (MAIN), a sound source driver processing program (TG_DRV), and a song processing program (SG_
PLY) is set. That is, FIG.
In the case of the memory capacity shown in (d), the main routine program (MAIN), the sound source driver processing program (TG_DRV), and the song processing program (SG_P)
LY) is copied to the DW-RAM3. Furthermore,
When the number of DRAMs constituting the DW-RAM 3 is increased so much that the memory capacity shown in FIG. 2F is increased, a DELAY area used as delay means,
U11 very large work area (WORK)
Area and the copied main routine program (MA
IN) and sound source driver processing program (TG_DRV)
An area for storing a song processing program (SG_PLY) and a style processing program (SL_PLY) is set. In other words, when the memory capacity is as shown in FIG. 2D, the main routine program (MAIN), the sound source driver processing program (TG_DRV), the song processing program (SG_PLY), and the style processing program (SL_PLY) are transferred to the DW-RAM 3. Be copied.

When the program is copied from the WP-ROM 2 to the DW-RAM 3 as shown in FIGS. 2B to 2F, the copied program is transferred to the DW-R
The pointer information indicating the start address of the area where the copied program is stored is written as the pointer information of the corresponding program in the control information table shown in FIG. As shown in FIG. 3, the control information table includes TG_DRV pointer information indicating a head address of an area in which a sound source driver processing program (TG_DRV) is stored,
SG_PLY pointer information indicating the start address of the area where the song processing program (SG_PLY) is stored;
SL_PLY pointer information indicating the start address of the area where the style processing program (SL_PLY) is stored, pointer information of other pointers required when executing the program, and the start of the area where the work area for the CPU 11 is set WORK area information indicating an address and DELAY area information indicating a head address of an area in which an effect delay area is set are stored. Further, the main routine program (MAIN) is written to the WP-ROM 2 from its start address, and when the main routine program (MAIN) is copied to the DW-RAM 3,
Since the data is written to the RAM 3 from the position of the head address,
Storing the pointer information in the register can be omitted.

Next, a start-up routine executed first when the power of the tone generator of the present invention is turned on will be described with reference to a flowchart shown in FIG. When the startup routine is started, in step S11, the DW-RA
M3 is cleared, its capacity is checked, and its capacity information is stored in the register YJ. This capacity information has five levels of “0” to “4”, and the capacity information indicating that the state shown in FIG. 2B is very small is “0”, and the state shown in FIG. The capacity information indicating that the state is small is “1”, the capacity information substantially indicating the middle state of the state illustrated in FIG. 2D is “2”, and the capacity information indicating the state of the large state illustrated in FIG. 3 ", the capacity information indicating that the state shown in FIG. 2 (f) is very large is" 4 ". For example, if the capacity of the DW-RAM 3 is 256 kbits or less (b), if the capacity of 257 kbits to 512 kbits (c),
If it is 513 kbit to 1024 kbit, it is determined as (e), and if it is more than that, it is determined as (f). Next, in step S12, it is determined whether or not the value of the register YJ is "0". If the value of the register YJ is "0", the process branches to step S21 to set the area set in the DW-RAM 3 in this case. Is stored in the control information table shown in FIG. In particular,
In this case, since the DW-RAM 3 has a very small capacity, the CPU 11 sets the DW-RAM 3 in accordance with the capacity.
Is divided into a DELAY area and a work area area, and the head address of each area is stored as WORK area information and DELAY area information.

When the process in step S21 is completed, a main routine program described later stored in the WP-ROM 2 is started and executed in step S22, and the startup routine is ended. When the value of the register YJ is "1"
If the above has been determined, the determination in step S12 is NO, and the process proceeds to step S13, where the highest priority main routine program is read from the WP-ROM 2 and the DW-R
The data is transferred to AM3 and written from the start address. Next, in step S14, the program is read from the WP-ROM 2 in accordance with the value of the
A determination process for transferring to M3 is performed. If it is determined in step S14 that the register YJ is "1", the memory capacity is small, and the process proceeds to step S15 in which no transfer is performed except for the main routine program. If it is determined in step S14 that the register YJ is "2", the memory capacity is medium. Therefore, in addition to the top-priority main routine program, the next highest priority tone generator driver processing program (TG_DRV) is stored in the WP-ROM2. Read and D
The process proceeds to step S16 for transferring to the W-RAM3. Also in this case, the main routine program is written from the head address of the DW-RAM3.

Further, at step S14, the register YJ
Is determined to be "3", the memory capacity is large.
In addition to the main routine program, a sound source driver processing program (TG_DRV) and a song processing program (S
G_PLY) from the WP-ROM 2 and read DW-
The process proceeds to step S17 for transferring to the RAM3. Also in this case, the main routine program is written from the head address of the DW-RAM3. Furthermore, if the register YJ is determined to be "4" in step S14, the memory capacity is very large, so that in addition to the main routine program, the tone generator driver processing program (TG_DRV), the song processing program (SG_PLY) and the style processing The process proceeds to step S18 where the program (SL_PLY) is read from the WP-ROM 2 and transferred to the DW-RAM 3. Also in this case, the main routine program is written from the head address of the DW-RAM3. Then, step S1
When the processing of Steps 5 to S18 is completed, information of the area set in the DW-RAM 3 in Step S19 is stored in the control information table shown in FIG.

In this case, the DEW is stored in the DW-RAM3.
An area in which at least a main routine program is stored is set in addition to the Y area and the work area area.
However, the main routine program is always DW-RAM
3, the pointer information is omitted, the head addresses of the DELAY area and the work area area are automatically allocated and stored in the control information table as WORK area information and DELAY area information. Further, TG_DRV pointer information, SG_
As the PLY pointer information and SL_PLY pointer information, the head address of the program copied to the DW-RAM 3 is stored in the control information table. For programs that are not copied to the DW-RAM 3, significant pointer information is not stored in the control information table. When the processing in step S19 ends, a main routine program described later copied to the DW-RAM 3 in step S20 is activated, and the activation routine ends.

Next, the main routine started in step S20 or step S22 of the start routine will be described with reference to the flowchart shown in FIG. WP
When the main routine is started by reading the routine from the top address of the ROM 2 or the top address of the DW-RAM 3, at step S31, various registers are cleared, and processing for preparing a screen to be displayed on the panel display is performed. Initial setting processing is executed. Then
In step S32, it is checked whether or not there is an activation factor. The activation factors include (1) a performance event being input or generated, (2) an operation event for controlling automatic performance and a timer interrupt event for automatic performance, and (3) an operation event for controlling automatic accompaniment. And an automatic accompaniment timer interrupt event, (4) a panel switch operation event, and (5) an end switch operation event.

Therefore, in step S32, it is checked whether or not any one of the above-mentioned five causes is present, and then, in step S33, it is determined whether or not the cause is present. When it is determined that there is a start factor, the process proceeds to step S34, and when no start factor is detected, the process returns to step S32 to wait for occurrence of the start factor. In step S34, when the activation factor (1) is detected, a sound source driver process is executed in step S35, and the process returns to step S32. At this time, the sound source driver processing program is DW-R
If it is copied to AM3 (the value of the register YJ is "2" or more), a significant TG_DRV pointer is written in the control information table shown in FIG. 3, so the routine is read from the address of the pointer and executed. Thus, the sound source driver processing program copied to the DW-RAM 3 can be executed. DW-
If not copied to the RAM 3 (the value of the register YJ is “1” or less), the sound source driver processing program is read from the WP-ROM 2 and executed.

In this tone generator driver process, tone generator parameters corresponding to the input event are set in the tone generator register 20, and a preparation process for generating a musical tone is performed. For example, when a note-on event is detected from the keyboard 9 or the MIDI interface 13, the sound source register 20
, Parameters such as note number, timbre information (waveform address), velocity (envelope parameter), modulation parameter, and the like are set in the channel area assigned by. When an event for changing the envelope shape is input, the changed envelope parameters are set in the tone generator register 20.

Thus, in the tone generation processing performed in the waveform memory tone generator when the tone generation timing is reached, the address reading circuit 21 is set in the tone generator register 20 in accordance with the time division channel timing obtained by equally dividing one sampling period. WP based on the tone color information of each channel and the read address corresponding to the note number.
-The waveform sample data is read from the ROM 2. The read waveform sample data is provided with an envelope in accordance with the velocity and timbre information in the EG providing unit 22, and further, as described above, the DSP 2 with mixer is provided.
In 3, the waveform sample data of each channel is mixed in a plurality of ways, and an effect is given to a part of the data. The digitized analog waveform signal from the analog waveform input unit 4 is also mixed in the DSP with mixer 23. The tone data thus generated is read out at a predetermined sampling cycle, and
The sound is generated from the sound system 7 via C6.

When the factor (2) is detected, the process proceeds from step S34 to step S36, where song processing is performed. At this time, the song processing program (SG
_PLY) is copied to the DW-RAM 3 (the value of the register YJ is “3” or more), the DW-RAM 3 indicated by the SG_PLY pointer in the control information table
The above-mentioned song processing program is used. Otherwise, the song processing program on the WP-ROM 2 is used. In this song processing, automatic performance processing for generating a performance event at the timing according to the musical score is performed in accordance with the music data read from the automatic performance file.
Return to 32. Note that an event generated according to the activation factor (2) is also a factor of the activation factor (1). Further, when the activation factor (3) is detected, the process proceeds from step S34 to step S37, where style processing is performed. At this time, the style processing program (SL-PLY)
If the data has been copied to the RAM 3 (the value of the register YJ is “4”), the style processing program on the DW-RAM 3 indicated by the SL_PLY pointer in the control information table is used; otherwise, the WP-ROM 2 Use the style processing program above. In this style processing, processing of generating a performance event at the timing of a musical score according to the accompaniment data of the set style and the like is performed,
It returns to step S32. An event generated in response to the activation factor (3) also becomes an activation event (1) as an input event.

Further, when the activation factor (4) is detected, the process proceeds from step S34 to step S38, where other processes are performed and the process returns to step S32. Other processes include a part volume control process which is performed when a part-specific volume setting is performed by operating an operator for setting a volume for each part provided on a panel switch. Furthermore, when the activation factor (5) is detected,
Proceeding from step S34 to step S39, predetermined termination processing for terminating a program relating to tone generation in the tone generator is performed, and the main routine is terminated. It should be noted that in the activation factors (4) and (5), the WP
-The other processing program (OTHER) and the termination program on the ROM 2 are used.

In the above description, the DW-RAM 3 is
The present invention is not limited to this, but the present invention is not limited to this. For example, the DW-RAM 3 may be any write / read memory with a high access time. It may be a memory. In the above-described embodiment of the present invention, the tone generator driver processing program (TG_DRV) is copied second after the main routine program. The event routine is also processed by the sound source driver processing program,
This is because the processing of the automatic performance and the automatic accompaniment can be made more efficient by copying the sound source driver processing program.

Furthermore, in the embodiment of the present invention, the other processing programs (OTHER) and the program of the end processing are not copied even if the DW-RAM 3 has a margin, because those programs are not used for the performance, This is because sex is not required. Instead of storing the pointer indicating the start address of the area where the copied program is stored in the control information table, a pointer indicating the start address of the area where the program to be used is stored may be stored. . In this case, if there is a program copied to the DW-RAM 3 for each routine, a pointer indicating the area is stored, and if not, a pointer indicating the area of the program in the WP-ROM 2 is stored.

[0037]

As described above, according to the present invention, when there is a margin in the capacity of an expandable high-speed memory such as a DRAM,
Since a high-priority program among a plurality of programs stored in a low-speed memory such as a ROM is copied to a high-speed memory, the CPU reads the program from the high-speed memory for the copied program. Can be performed. Therefore, since the program can be executed at high speed, the performance of the CPU can be improved. In this case, the CPU accesses the high-speed memory, and the sound source unit accesses the ROM in which the waveform sample data is stored.
Since the contention of memory access to the same memory does not occur between the PU and the tone generator, the performance of the CPU can be prevented from deteriorating, and the CPU can perform high-speed arithmetic processing.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration example of an embodiment of a tone generator according to the present invention.

FIG. 2 is a diagram showing an example of a memory map in the musical sound generator of the present invention.

FIG. 3 is a diagram showing an example of a control information table in the musical sound generator of the present invention.

FIG. 4 is a diagram showing a flowchart of a startup routine in the musical sound generating device of the present invention.

FIG. 5 is a diagram showing a flowchart of a main routine in the musical sound generating device of the present invention.

[Explanation of symbols]

1 sound source LSI, 2 WAVE & PROGRAM RO
M, 3 DELAY & WORK RAM, 4 analog waveform input, 5 ADC, 6 DAC, 7 sound system, 8 MIDI terminal, 9 keyboard, 10 panel / external CPU, 11 CPU, 12 timer, 13 MIDI
Interface, 14 serial I / O, 15 parallel I / O, 16 access management A, 17 access management B, 18 bus line, 20 sound source register, 21 readout circuit, 22 EG adding unit, 23 DSP with mixer

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI G06F 17/10 H03H 17/02 601H H03H 17/02 601 681H 681 G06F 9/06 420C 420H 221066 (JP, A) JP-A-6-222276 (JP, A) JP-A-57-89144 (JP, A) JP-A-10-21082 (JP, A) JP-A-60-129855 (JP, A) JP-A-60-204029 (JP, A) JP-A-1-105611 (JP, A) JP-A-5-275973 (JP, A) JP-B-61-38496 (JP, B2) (58) (Int.Cl. ⁷ , DB name) G10H 1/00-7/12 G06F 9/445 G06F 12/08 H03H 17/02 681 G06F 17/10

Claims (2)

(57) [Claims] A read-only memory shared on a memory bus stores a plurality of programs read and executed by an arithmetic processing unit and waveform sample data read by a waveform memory sound source unit and generating musical tone waveforms. And the waveform memory sound source section is provided from the read-only memory.
The waveform sample data is read out to generate a musical tone waveform, the arithmetic processing unit detects a high-speed memory capacity in which a work area area used when executing the program is set, and detects the detected high-speed memory capacity. Is copied from the read-only memory to the high-speed memory, and when the program is executed, the arithmetic processing unit copies the program copied to the high-speed memory to the waveform memory sound source unit. Is the read-only
The buffer used when reading waveform sample data from memory.
A tone generation method, wherein the tone is read out and executed using a bus different from the bus . 2. An arithmetic processing unit, a waveform memory tone generator for generating musical sound data, a plurality of programs executable by the arithmetic processing unit, and waveform sample data read by the waveform memory tone generator when generating a musical tone waveform are stored. A read-only memory shared on a memory bus by the arithmetic processing unit and the waveform memory sound source unit, wherein the waveform memory sound source unit is
The waveform sample data is read out to generate a musical tone waveform, the capacity of a high-speed memory in which a work area for the arithmetic processing unit is set at startup is detected, and the number of programs corresponding to the detected high-speed memory capacity is detected. the running process of copying from the read only memory to said high speed memory, when executing the program, a program that the processing unit is copied to the high-speed memory, the
The waveform memory sound source section receives the waveform data from the read-only memory.
A bus different from the bus used to read sample data
Musical tone generating apparatus being characterized in that so as to perform reading using the scan.