JP4106798B2 - Sound generator - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a sound source device that assigns a sound generation slot to generate sound for a plurality of input sound generation instructions.
[0002]
[Prior art]
The sound source device can simultaneously generate a plurality of musical sounds such as 32 sounds and 64 sounds, but depending on the form of performance, there are cases where more musical sounds are generated simultaneously. In order to cope with such a case, the conventional tone generator has a function called a key assigner, and in response to a new pronunciation instruction, one of the musical sounds currently being produced is selected and forcibly muted (force dump) ).
[0003]
Which sound is to be force dumped is a very serious problem in the performance of music, and if it is to be judged accurately, a very complicated judgment is required. For this reason, conventionally, since the sound generation slot to be force-dumped is determined by simple judgment such as the sound generation duration time, there is a problem that an important conspicuous sound is erased.
[0004]
An object of the present invention is to provide a sound source device that realizes an accurate force dump by quickly searching for an inconspicuous sound.
[0005]
[Means for Solving the Problems]
According to the first aspect of the present invention, a plurality of sound generation slots that independently form musical sounds, and in response to a sound generation instruction, the sound generation slot is instructed to form a tone of a timbre according to the sound generation instruction, and each sound generation slot Control means for instructing the stop of the formation of the musical sound when the amplitude level of the musical sound waveform formed by is less than or equal to the first threshold value, and the control means newly adds the musical sound during formation of the musical sound in all the sound generation slots. If there is a pronunciation slot that produces a rhythm sound, and if there is a pronunciation slot that produces a rhythm sound, the amplitude level of the sound waveform formed by each pronunciation slot when the output value is a value obtained by multiplying the gain of the formed tone waveform is further searched pronunciation slot is a second threshold value or less, if there is a corresponding sound slots, this sound slot Comprising a silencer slot determination means for determining a pronunciation slots to force mute, characterized by instructing the formation of a musical tone in accordance with the new sound generation instruction to force mute the sound slot.
[0006]
According to a second aspect of the present invention, in the first aspect of the invention, the mute slot determining means determines a sound generation slot whose musical sound waveform is in a release stage when there is no sound generation slot of a rhythm sound whose output value is equal to or less than a second threshold value. If there is a sound generation slot that is in the release stage, a sound generation slot whose output value is less than or equal to the second threshold value is searched for. If there is a sound generation slot, the sound generation slot is forcibly muted. It is a means to determine as a pronunciation slot.
[0009]
The sound source device of the present invention includes a plurality of (32, 64, 128, etc.) sound generation slots and can simultaneously form a plurality of musical sounds. However, since the pronunciation instruction is not always input within the range of the maximum number of pronunciations, when a pronunciation instruction is input exceeding the maximum number of pronunciations, the musical sound of one of the pronunciation slots is force dumped. The sound generation slot is released by this, and a new musical sound is generated in this sound generation slot. Which musical sound is to be force dumped can be determined by selecting the least important musical sound from each sounding slot, but the importance is most determined based on the level of that musical sound. Easy and reliable.
[0010]
Therefore, in the present invention, the musical sound to be force dumped is determined based on the amplitude level (EG) at the time of forming the musical sound waveform and the gain (VOL) at the time of outputting the musical sound. As a result, it is possible to search for a musical tone having a volume that can be silenced among many musical sounds that are being generated. In addition, some musical sounds have gains that increase or decrease, such as fading out and then fading in again. By selecting an appropriate tone such as a percussion instrument sound or an appropriate envelope stage such as during release, this tone can be selected. Such a musical sound can be removed, and a reliable selection becomes possible.
[0011]
Further, the processing is further simplified by determining whether or not the output level calculated by EG × VOL in each sound generation slot is equal to or less than the threshold value.
[0012]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a block diagram of a sound source device according to an embodiment of the present invention. This sound source device has 32 sound generation slots and can simultaneously form 32 musical sounds. On the other hand, when a new (33rd sound) sound generation instruction is generated while all sound generation slots are operating, the least important sound is selected from the sound being sounded, and this is force dumped (forced mute). Then, a new musical tone is formed in this pronunciation slot.
[0013]
The sound generation slot to be muted is selected by selecting an EG level that is an amplitude change curve value of a musical tone, a VOL level that is a gain of amplitude control such as fade-in / out, voice (instrument tone), and status (EG stage: attack, decay) , Sustain, release, etc.).
[0014]
Referring to FIG. 1, the tone generator includes a CPU 1 that controls the operation of the entire apparatus, a tone generator LSI 2 that synthesizes musical sound waveforms to generate musical sounds, a memory 3 that stores programs and data, and a MIDI message that is input from the outside. The MIDI buffer 4 to be connected, the MIDI interface 5 connected to the MIDI buffer 4, the operation / display panel 7 for displaying the setting contents and directly performing the setting operation without using the MIDI, and the operation / display panel 7 are connected to the CPU 1. The panel interface 6, the effector 8 that gives an effect to the musical sound formed by the sound source LSI 2, and the D / A converter 9 that converts the musical sound that the effector 8 gives the effect into an analog signal.
[0015]
The tone generator LSI 2 is also provided with an input buffer 10 for the CPU 1 to write parameters and an output buffer 11 for notifying the CPU 1 of the internal operation status. The CPU 1 and the tone generator LSI 2 exchange information via the input register 10 and the output register 11.
[0016]
FIG. 2 is a configuration diagram of the sound source LSI 2. The tone generator LSI 2 has 64 sound generation slots 21 and can simultaneously generate 64 musical sounds. The operation of each sound generation slot is controlled by the CPU 20 in the sound source LSI. The CPU 20 controls each sound generation slot 21 based on the data written to the input register 10 and periodically writes the operation state of each sound generation slot 21 to the output register 11.
[0017]
The sound generation slot 21 has a waveform forming unit 22 and an amplifier 23. The waveform forming unit 22 forms a musical sound waveform with digital data of a predetermined sampling rate based on data input from the CPU 20. The data supplied to the waveform forming unit 22 includes voice (instrument timbre), that is, data for determining the shape of a musical sound waveform, data for generating a waveform amplitude change curve (EG), parameters for designating the frequency of the musical sound, and the like. It is. The data input to the amplifier 23 is a parameter for determining the gain VOL for musical tone amplification.
The musical sound formed by each sound generation slot is mixed into a 2-channel stereo signal by a mixer and then output externally.
[0018]
Further, the CPU 20 of the tone generator LSI 2 periodically writes the operation status of each sound generation slot in the output register 11. The written content is read by the CPU 1. In the output register 11, the note on / off status (whether or not a musical tone is formed) of each sound generation channel is first written, and information such as the EG level and the VOL level is written for the sound generation channel in which the note is on. It is.
[0019]
The EG level is a value indicating a temporal change in the amplitude level of the musical tone waveform to be formed, and is generally a curve having a shape as shown in FIG. This EG curve is divided into four stages, where a is an attack stage, b is a decay stage, c is a sustain stage, and d is a release stage. This EG is a parameter that controls the formation of the tone waveform itself. The VOL level is a parameter that determines how much the musical sound is amplified at the output stage of the formed musical sound. Therefore, the musical sound output from the sound source is output at an output level that is a value obtained by multiplying these (EG × VOL), as shown in FIG. 7B.
[0020]
In addition to an automatic performance device such as a sequencer, the MIDI interface 5 is connected to a performance device for electronic musical instruments such as a keyboard and a wind instrument type controller. The automatic performance device is a device for automatically playing music by outputting MIDI messages at a predetermined timing by sequentially reading out automatic performance data stored in advance. The performance device is operated by the performer, and generates and outputs a MIDI message in accordance with the operation. MIDI messages output from these devices are written to the MIDI buffer 4 via the MIDI interface 5.
[0021]
MIDI messages include, for example, a note-on message that instructs the sound to be played, a note-off message that instructs the mute, a program change message that specifies the timbre (voice), a control change message that controls the timbre (waveform) and effects, etc. is there.
[0022]
The CPU 1 periodically reads the contents of the MIDI buffer 4 and executes processing according to the contents of the written MIDI message. For example, when a note-on message is input, the sound source LSI 2 is instructed to form a musical sound corresponding to the note-on message.
[0023]
This musical tone formation (pronunciation) instruction is performed in the following procedure. First, it is determined whether the tone generator LSI 2 has a sound generation slot (empty slot) that is not currently operating. In some cases, one of them is assigned to this musical tone, and the necessary parameters are written in the area corresponding to the tone generation slot of the input buffer 10. This parameter is a note-on instruction, a voice instruction, touch intensity data, or the like. The voice is a voice designated as the MIDI channel of the note-on message, and is stored in the MIDI channel information table set in the memory 3.
[0024]
FIG. 3 is a configuration diagram of the memory 3. The memory 3 includes a ROM 3a, a battery-backed SRAM 3b, and a DRAM 3c.
The ROM 3a stores a control program (see flowcharts in FIGS. 4 to 6) for controlling the operation of the sound source device, default parameters set in each part of the sound source device in the initial state, and a plurality of voices (instrument timbres). Voice waveform data, which is data necessary for forming a musical sound waveform, is stored in advance. The SRAM stores user setting parameters, user-designed voice waveform data, and the like, and is backed up even when the power is turned off. The voice waveform data includes data for determining the shape of the waveform, data for forming an EG waveform that is a temporal variation curve of the amplitude value, and the like.
[0025]
Furthermore, various storage areas used by the control program for managing the operation of the tone generator LSI 2 are set in the DRAM. Sound generation flag Fg (i) indicating whether the musical sound of each sound generation slot is in a note-on state (i = 1 to 64: the same applies hereinafter), an active flag Fa indicating whether each sound generation slot is forming a musical sound waveform ( i) Sound generation information table G (i) for storing information such as MIDI channels, voices, pitches, sound generation start times, etc. of musical sounds formed by each sound generation slot, EG levels of musical sounds formed by each sound generation slot EG buffer EG (i) for storing the sound, output level buffer Lo (i) for storing the output level of the musical sound formed by each sound generation slot, MIDI channel information table for storing the setting contents of each MIDI channel, etc. are set. The
[0026]
Note that even if a note-off message is input, the musical sound is not immediately muted, but the sounding flag that indicates the note-on / off state and the active flag that indicates whether or not the musical tone is being formed are separate because the sound is gradually attenuated and then muted. Is provided. In the storage area, the EG buffer and the output level buffer are sequentially updated by the sound source management process shown in FIG.
[0027]
The operation of this sound source device will be described with reference to the flowcharts of FIGS.
FIG. 4 is a flowchart showing the sound source management process. This process is an operation in which the CPU 1 checks the operation status of each sound generation slot of the tone generator LSI 2 and is periodically executed. First, 1 is written in the pointer i indicating the number of the sound generation slot (s1). Then, it is determined whether or not the active flag Fa (i) of the sound generation slot indicated by i is set (s2). If the active flag Fa (i) is not set, this tone generation slot is an empty slot because no musical tone is formed, and it is not necessary to check the status of this slot, so the process jumps directly to s9.
[0028]
When Fa (i) is set and the sound generation slot designated by i is in operation, the EG level value of the musical sound waveform currently formed from the area corresponding to this sound generation slot i of the output register 11 is set. Reading (s3), it is determined whether this is below the first threshold value Tha (s4).
[0029]
The first threshold value Tha is a threshold value indicating a level that is so small that it cannot be heard even if this musical tone is output alone. If the amplitude level of the musical tone is lower than this, the note-off message is sent from the MIDI interface. Even if no is input, the value is such that it can be automatically determined that the note-off / musical tone formation has ended.
[0030]
When the EG level value exceeds the first threshold value Tha, this value is written as the current EG level value in the storage area EG (i) corresponding to the sound generation slot i of the EG buffer (s5).
[0031]
Next, the VOL level value of this tone generation slot i is read from the output register 11 of the tone generator LSI 2 (s6). This VOL level value may be controlled by any of parameters applied from the outside via the MIDI interface 4, parameters read / generated by the CPU 1, and parameters read / generated by the tone generator LSI 2. The read VOL level value is multiplied by the EG level value to calculate an output level which is the level of the musical tone actually output from the sound source (s7), and this value corresponds to the sound generation slot i of the output level buffer. The current output level is written in the storage area Lo (i) (s8).
[0032]
The values of these EG (i) and Lo (i) are referred to when a new note-on message is input or during force dump (forced mute processing), which is processing when there is no empty slot in the sound source LSI. .
[0033]
On the other hand, if the EG level value is equal to or lower than the first threshold value Tha in s4, the tone generation slot assignment canceling process is executed. First, the assignment cancellation is written in the area of the tone generation slot i of the input register 10, and the tone generation operation of this tone generation slot is terminated and released (s11). Then, the active flag Fa (i) is reset (s12), and the EG buffer EG (i), the output level buffer Lo (i), and the sound generation table G (i) are cleared (s13). Further, if the note-off message has not yet been input for this musical sound and the sound generation flag remains set (s14), the sound generation flag Fg (i) is reset (s15).
[0034]
After the above processing is repeatedly executed for each sound generation slot with i = 1 to 64 (s9, s10), the operation is terminated. The number of sound generation slots of the tone generator LSI 2 is not limited to 64.
[0035]
The flowchart of FIG. 5 shows a MIDI process for fetching a MIDI message from the MIDI buffer 4 and executing the corresponding process. First, the contents of the MIDI buffer 4 for buffering external MIDI input are read (s21). It is determined whether or not the stored content is a note-on message (s22). If it is not a note-on message, the corresponding processing (s31 and subsequent steps) is performed.
[0036]
When a note-on message is input, the following processing is executed to cause the tone generator LSI 2 to generate a musical sound corresponding to the message. First, it is determined whether or not there is an empty sounding slot (s23). This may be determined based on whether there is a sound generation slot in which the active flag Fa (i) is reset. If there is an empty slot, one of the empty slots is assigned as a sounding slot corresponding to the current note-on message, and is necessary for forming a musical sound waveform in the area corresponding to this sounding slot of the input register 10 A new parameter is written (s24). By writing the necessary parameters in the input register 10, the tone generator LSI 2 starts to generate a musical tone in this tone generation slot. Thereafter, the sound generation flag Fg (i) and the active flag Fa (i) corresponding to the sound generation slot i are set (s25), and the sound generation information such as the voice and the sound generation start time corresponds to the sound generation slot i of the sound generation information table. Store in area G (i) (s26). Then, the maximum value is written in the storage areas EG (i) and Lo (i) corresponding to the sound generation slot i of the EG buffer and output level buffer (s27), and the assignment process is terminated.
[0037]
The maximum value of the EG level and the output level is a maximum value (127) in a range of values (for example, 0 to 127) that can be taken by the parameter indicating the EG level and the output level. The actual sound waveform formation has not yet started in the sound generation slot immediately after the start of sound generation, but the maximum value of the EG level is written to increase the priority of the sound generation slot so that force dumping is not performed.
[0038]
On the other hand, when it is determined in s23 that there is no empty slot, that is, when all the sound generation flags are set, it is determined which sound sound of the sound generation slot is force dumped (forced mute). (S28). This force dump determination rule will be described later.
[0039]
Then, the sound source LSI 2 is instructed to perform a force dump of the sound generation slot i determined to be silenced (s29). This instruction is performed by writing a force dump instruction in an area corresponding to the sound generation slot i of the input register 10. When a force dump is instructed, the sound generation slot abruptly attenuates the level of the tone being formed and releases the slot almost immediately. Then, data for forming a new musical tone is written in the input register 10 of the sound generation slot that has been force dumped to start the musical tone formation (s30). The sound generation flag and active flag remain set before the force dump, so there is no need to set them again. Then, sound generation information such as a tone color assigned to the sound generation slot i is stored in the sound generation information table G (i) (s26), and the maximum value is written in the EG buffer EG (i) and the output level buffer Lo (i) ( s27), the assignment process is terminated.
[0040]
If the message input to the MIDI buffer is a note-off message (s31), the sound generation information table is searched for which sound generation slot corresponds to the note-off message (s32). This can be done by searching for the MIDI channel and pitch of the note-off message. If there is a corresponding sounding slot (s33), a note-off instruction is given to the sounding slot (s34), and the corresponding sounding flag Fg (i) is reset (s35).
[0041]
If the corresponding sound generation slot is not found in the search of s32, it is determined that the sound muting process has already been performed automatically based on the EG level or the force dump process has been performed. To do.
[0042]
If the MIDI message input to the MIDI buffer 4 is something other than a note-on message and a note-off message, a process corresponding to the message is executed (s36), and the process ends.
[0043]
When the input MIDI message is, for example, a control change message for changing the VOL value of a certain MIDI channel, the sound generation slot that generates the musical sound of the MIDI channel is searched from the sound generation information table, and the input register 10 A process of writing a new VOL value in the area corresponding to the sound generation slot is performed. In the tone generator LSI 2, the CPU reads this VOL value and supplies it to the corresponding sound generation slot.
[0044]
FIG. 6 is a flowchart showing a process for determining a sound generation slot to be force dumped. First, the voice of the sound that is sounded in each sound generation slot is read out from the sound generation information table, and a search is made as to whether there is a sound generation slot that generates a rhythm sound (s40). If there is a tone color of the rhythm sound, it is searched whether there is an output level buffer Lo (i), that is, a value of EG × VOL equal to or smaller than the second threshold Thb (s41). If there is a corresponding sound generation slot, this sound generation slot is determined as a sound generation slot for force dump (s45). If a plurality of sound generation slots are applicable, the sound generation slot may be selected by selecting the one with the lowest EG level and the longest sound generation time.
[0045]
The rhythm sound is searched for as the first candidate for the force dump because the rhythm sound is a musical sound whose EG level is monotonously attenuated, and the VOL is also attenuated once and then hardly faded in again. That's why. In the case of a rhythm sound, note-off messages are often not input, and in this case, the sound generation slot may not be released even if the sound is attenuated to such an extent that it cannot be heard.
[0046]
As described above, the second threshold value Thb is a threshold value for the output level (EG × VOL). When all sound generation slots are operating and 64 sounds are generated, the second threshold value Thb is almost audible at this output level. This is a threshold value indicating the level of no. That is, although it is not a sound that cannot be heard, it indicates that the sound can be muted if all sound generation slots are sounding.
[0047]
On the other hand, if there is no rhythm sound whose output level Lo (i) is equal to or lower than the second threshold value Thb, the currently released musical sound is searched (s42). The musical tone currently being released is one in which the note off message is received and the sound generation flag Fg (i) is reset, but the EG status is still released and the active flag Fa (i) is set. Among them, a search is made for an output level Lo (i) that is equal to or lower than the second threshold Thb (s43). If there is a corresponding sounding slot, this sounding slot is determined as a sounding slot for force dump (s45). If a plurality of sound generation slots are applicable, the sound generation slot may be selected by selecting the one with the lowest EG level and the longest sound generation time.
[0048]
The musical tone being released only attenuates monotonously, and the EG level does not rise again, and the VOL does not fade in. Therefore, it is possible to force dump as long as it is below the second threshold Thb. This is because it can be determined.
[0049]
Further, if there is no released musical tone whose output level Lo (i) is attenuated to the second threshold value Thb or less, the tone generation slot having the smallest value of EG (i) is selected from the musical tone being generated. Search is performed (s44), and this is determined as a sound generation slot for force dump (s45).
[0050]
Thus, even when a note-on message exceeding the maximum number of pronunciations of the tone generator LSI is input, the music that is being played is selected by efficiently selecting the music that has the least influence and force dumping the music. Never break the atmosphere.
[0051]
Thus, in this embodiment, the first threshold value for determining mute based on the EG level and the second threshold value for permitting force dump based on the output level (EG × VOL) are separately set. Provided.
[0052]
【The invention's effect】
According to the present invention, by determining which sound generation slot is to be forcibly silenced based on the amplitude level and gain, it is possible to accurately find the musical sound with the least influence with simple judgment. .
[Brief description of the drawings]
FIG. 1 is a block diagram of a tone generator according to an embodiment of the present invention. FIG. 2 is a block diagram of a tone generator LSI of the tone generator. FIG. 3 is a memory block diagram of the tone generator. FIG. 5 is a flowchart showing the operation of the sound source device. FIG. 6 is a flowchart showing the operation of the sound source device. FIG. 7 is a diagram showing an example of changes in the EG parameter and the VOL parameter of the sound source device. ]
1 ... CPU, 2 ... sound source LSI, 3 ... memory, 4 ... MIDI buffer, 5 ... MIDI interface,
DESCRIPTION OF SYMBOLS 10 ... Input register, 11 ... Output register, 20 ... CPU (sound source LSI), 21 ... Sound generation slot, 22 ... Waveform forming unit, 23 ... Amplifier, 24 ... Mixer

Claims (2)

Multiple sound generation slots that each independently form a musical tone,
In response to a sound generation instruction, the sound generation slot is instructed to form a tone color tone corresponding to the sound generation instruction, and a musical sound waveform whose amplitude level is less than or equal to a first threshold is formed. Control means for instructing to stop the formation of
With
The control means searches for a pronunciation slot that is producing a rhythm sound when there is a new pronunciation instruction while forming a musical sound in all the pronunciation slots, and if there is a pronunciation slot that produces a rhythm sound, Among them, a sound generation slot whose output value, which is a value obtained by multiplying the amplitude level of the sound waveform formed by each sound generation slot by the gain of the formed sound waveform, is equal to or less than the second threshold is further searched. When there is a corresponding sound generation slot, the sound generation slot is provided with a mute slot determination means for determining the sound generation slot as a sound generation slot for forced mute,
A sound source device that instructs the formation of a musical sound in response to the new sound generation instruction to the sound generation slot forcibly muted. The mute slot determining means searches for a sound generation slot whose musical sound waveform is in the release stage when there is no sound generation slot of the rhythm sound whose output value is equal to or less than the second threshold, and if there is a sound generation slot in the release stage, 2. The sound generating slot having an output value equal to or smaller than a second threshold value is searched for, and when there is a corresponding sound generating slot, the sound generating slot is determined as a sound generating slot for forced mute. Sound source device.

Images