JP2001282241A - Electronic musical tone generator, electronic musical tone generating method and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an electronic musical tone generator which is capable of producing the plural musical tones to be simultaneously produced simultaneously or during a short time. SOLUTION: The electronic musical tone generator has an input means for inputting key-on information, a timer for starting the measurement of the prescribed time from the input of the key-on information when the key-on information is inputted, a parameter setting means for setting the parameters necessary for the sound production when the key-in information is inputted and a sound production start instruction means which instructs the sound production start to sound sources set with the parameters after lapse of the prescribed time from the input of the key-on information.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electronic musical tone generation technique, and more particularly to an electronic musical tone generation technique for an automatic performance (accompaniment) piano, a tone generator board with a built-in personal computer, a software tone generator for generating a waveform by a computer program, and the like.

[0002]

2. Description of the Related Art A technique for controlling the tone generation of electronic musical tones using a CPU has been applied for a patent from around 1974.
It has been commercialized since around the year. An advantage of using the CPU is that complicated processing can be executed at high speed depending on the computer program. Before the CPU is used for an electronic musical instrument, there is only a simple automatic accompaniment pattern, display on the instrument panel, assignment of a limited number of musical sound generation channels when simultaneously producing a plurality of output sequences (MIDI channels), and the like. Was also very simple. This CP
Utilizing the advantages of U, recent electronic musical sound generators
When one piece of key press information (key-on information) is received, an extremely large number of processes are performed to generate sound.

[0003] In the Showa era, panel processing, key assigner, automatic performance, and the like were the main processes of the CPU. At that time, for the timbre, the waveform read address, envelope parameter, and the like had to be read from the timbre coefficient memory and set as it was in the predetermined tone generation channel of the tone generator LSI. However, in recent years, after reading out these parameters, a key scale operation (that is, a process of storing only parameters corresponding to a specific key and obtaining other key-on information parameters by an interpolation operation or the like) or a velocity operation (attached to key-on) is performed. (Process of further changing the above parameters by key-on velocity data). That is why a real sound or a sound having a response to performance is required.

In addition, digital filters have become indispensable eras, and calculations are performed in which the coefficients of the digital filters change on the time axis depending on the timbre, pitch (key number), and key strength (velocity). Of course, it is also possible to leave too much computation to the sound source LSI.
Since it is desired to reduce the unit price of I, a load is often placed on the momentum CPU.

Here, CPUs used in electronic musical instruments
In consideration of the above, the processing capability has been improved by more than 100 times since starting with 8-bit CISC chips (8085, Z80, etc.) around 1982, and adopting 32-bit RISC chips in recent years. However, the time required for one key-on processing has not changed much between that time in 1982 and recent years. A long ago ago, I believed that the CPU processing capacity would increase in the future, so that variations in pronunciation due to simultaneous key depression could be ignored, but it seems that the atmosphere will be betrayed in the future.

This is because engineers have an implicit understanding that if the sound generation processing time is about 1 ms / key, there is no problem in hearing. In other words, no matter how much the processing capacity of the CPU is improved, it is likely that setting the processing capacity of 1 ms / key is the best in terms of cost performance. Therefore, in the near future, the sound source LSI will be only a plurality of multipliers and adders, and the input parameters will be supplied by the CPU almost in real time. In other words, if monaural, the sound source LS
I is becoming an unnecessary age.

[0007]

Here, there is a question whether the current processing time of the CPU is really a problem. Actually, about the piano sound (20
(a waveform requiring about ms), even if a plurality of sounds are activated at intervals of 1 ms, they can be heard simultaneously by the human ear at a general level. However, in the case where sounds of about eight keys are sounded at the same time with a sound with a sharper attack (such as a tibia sound or a percussion sound), it is likely that a person who recognizes the sound will appear.

The above-mentioned processing content of 1 ms / key is for the case where at most two tone generation channels (for L / R stereo) are used for one key-on. Output series (Piano L
/ R and L / R of strings)
Simply, one key requires a processing time of four tone generation channels (2 ms). In other words, it can be understood that the current CPU processing method is permissible when a normal person listens to the normal performance, but otherwise, the electronic musical tone generator is concerned about the sound generation delay.

FIG. 21 is a time chart showing a first example of a tone generation process of an electronic musical tone generator according to the prior art. The horizontal axis indicates the time from the key-on information input, and the vertical axis indicates the tone generation channel (hereinafter referred to as DCO) number. Simultaneously, three keys (keyA, keyB, keyC) are depressed, and a piano sound PN and a string sound ST are generated in stereo (L / R) for each key. In this case, two tone colors are used. Here, it is assumed that a processing time of 1 ms is required for one DCO.

A left (L) piano tone PN having a pitch A is assigned to the first DCO, and a right (R) piano tone PN having a pitch A is assigned to the second DCO. Are sounded in 2 ms from the key-on information input. Third DCO
Is assigned a left (L) string sound ST with a pitch A, and the fourth DCO is assigned a right (R) string sound ST with a pitch A, both of which are 4 ms from key-on information input. Pronounced as

The fifth DCO is assigned a left (L) piano tone PN having a pitch B, and the sixth DCO is assigned a right (R) piano tone PN having a pitch B. Are sounded in 6 ms from the key-on information input. Seventh DCO
Is assigned a left (L) strings sound ST with a pitch B, and an eighth DCO is assigned a strings sound ST with a right (R) pitch B, both of which are 8 ms from the key-on information input. Pronounced as

The ninth DCO is assigned a left (L) piano tone PN with a pitch C, and the tenth DCO is assigned a right (R) piano tone PN with a pitch C. Are sounded in 10 ms from the key-on information input. The eleventh DCO has a left (L) string sound ST with a pitch of C
Is assigned to the twelfth DCO, and the right (R) strings sound ST of the pitch C is assigned, and both are pronounced in 12 ms from the key-on information input.

In this case, since the first DCO produces a sound in 2 ms and the twelfth DCO produces a sound in 12 ms, it takes 10 seconds from the production of the first DCO to the production of the twelfth DCO.
It takes a long time of ms. Since the player presses all three keys at the same time, it is preferable that all the sounds are produced simultaneously. However, a plurality of musical tones are produced in 10 ms and the human ear is produced simultaneously. It is hard to feel that there is.

FIG. 22 is a time chart showing a second example of the tone generation process of the electronic musical tone generator according to the prior art. The horizontal axis shows the time from key-on information input, and the vertical axis shows DCO
Indicates a number. At the same time, 6 keys (keyA, keyB, key
yC, keyD, keyE, keyF) are depressed, and the tibia sound TA and the strings sound ST are generated in monaural for each key. Also in this case, two tone color series are used. Again, one meter for one DCO
It is assumed that the processing time of s is required.

A tibia sound TA having a pitch A is assigned to the first DCO, and is generated in 1 ms. Second DC
A string sound ST having a pitch A is assigned to O, and is generated in 2 ms. The third DCO has a pitch of B
Is assigned, and is generated in 3 ms. The fourth DCO has a string sound ST of pitch B
Is assigned and pronounced in 4 ms.

A tibia sound TA having a pitch C is assigned to the fifth DCO, and is generated in 5 ms. 6th DC
A string sound ST having a pitch C is assigned to O, and is generated in 6 ms. The seventh DCO has a pitch of D
Is assigned, and is generated in 7 ms. The eighth DCO has a string sound ST having a pitch D.
Is assigned and is pronounced in 8 ms.

The ninth DCO is assigned a tibia sound TA having a pitch E and is generated in 9 ms. Tenth D
A string sound ST having a pitch E is assigned to CO and is emitted in 10 ms. The tibia sound TA having a pitch of F is assigned to the eleventh DCO, and is generated in 11 ms. The twelfth DCO is assigned a strings sound ST having a pitch of F, and is sounded in 12 ms.

In this case, since the first DCO produces a sound in 1 ms and the twelfth DCO produces a sound in 12 ms, it takes 11 seconds from the first DCO to the twelfth DCO.
It takes a long time of ms. The player simultaneously presses the six keys, but a plurality of musical tones are generated in 11 ms, and it is difficult to feel that the tones are simultaneously generated.

An object of the present invention is to provide an electronic musical tone generating apparatus, an electronic musical tone generating method, and a recording medium capable of simultaneously producing a plurality of musical tones to be produced simultaneously or within a short time.

[0020]

According to one aspect of the present invention, an input means for inputting key-on information, and a timer for starting measurement of a predetermined time from the input of the key-on information when the key-on information is input. Parameter setting means for setting a parameter required for sounding when the key-on information is input to a sound source, and sounding start instructing means for instructing a sound source in which the parameter is set to start sounding after a predetermined time has elapsed from the input of the key-on information. Is provided.

According to another aspect of the present invention, (a) the step of setting parameters necessary for sound generation in the sound source based on the key-on information when the key-on information is input, and (b) the lapse of a predetermined time from the input of the key-on information Instructing the sound source in which the parameter is set to start sounding later.

According to yet another aspect of the present invention, (a)
A step of setting parameters required for sound generation in the sound source based on the key-on information when key-on information is input; and (b) a step of instructing a sound source in which the parameters are set to start sounding after a predetermined time has elapsed from the input of the key-on information. And a computer-readable recording medium on which a program for causing a computer to execute the program is recorded.

According to the present invention, the sound generation is not started immediately after the parameters required for sound generation are set in the sound source, but the parameter for a plurality of musical tones is waited until a predetermined time has elapsed from the input of the key-on information. Since the sound source is instructed to start sounding when set in the sound source, a plurality of musical tones can be sounded simultaneously or in a short time.

[0024]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below along with examples with reference to the drawings. In the present embodiment, when it is required to sound simultaneously with either one key or eight keys, sounding is started as simultaneously as possible. Therefore, a predetermined time (for example, from receiving key-on information to starting sounding) (for example, 6 ms) delay, during which necessary parameter calculations are completed. By doing this, for example, a person who plays an electronic piano may give a comment, "I wonder if the contact position of the key has become a little deeper", but from the beginning, only the electronic musical sound generator listens If it is not, no chords can be pronounced at the same time without any trouble.

The listener feels that the sound is being pronounced at the same time when the sound of a plurality of sounds is started within 10 ms or more from the experience of the inventor. If a plurality of sounds are generated within a shorter time of 4 ms or 6 ms, it sounds as if a plurality of sounds were generated at the same time. That is, if there is an interval of 10 ms or more from the start of the first sound generation to the start of the last sound generation, it does not seem to be received simultaneously.

In this embodiment, even if there is a margin for the processing, the key-on is intentionally delayed by a predetermined time (for example, 6 ms), and during that time, from the parameter calculation to the setting of the sound source to the corresponding tone generation channel is completed. And 6ms
Thereafter, for a plurality of tone generation channels for which parameters have already been set in the sound source, the sound generation is started by clearing the operation stop flags one after another. As a result, in this embodiment, the time from the start of sounding of the first tone generating channel to the start of sounding of the last tone generating channel can be shortened by several ms compared to the related art.

The intentional delay of the onset of sounding by the first key-on is evaluated with respect to automatic performance reproduction (including karaoke), automatic accompaniment reproduction, reproduction based on a key press operation of an electronic keyboard instrument, ensemble reproduction of an automatic performance piano, and the like. Then, only the electronic keyboard instrument pays attention to the delay time, but otherwise there is no problem in, for example, automatic performance and automatic accompaniment if the timing of reading out the key-on information is controlled.

Also, when electronic musical sounds are used for the ensemble sound of an automatic performance piano, the time required for the hammer of the raw piano to move after receiving the key-on information is orders of magnitude longer (slower). The current situation is that the onset of electronic musical sound is delayed lately, and there is no problem.

FIG. 1 is a time chart showing a first example of tone generation processing of the electronic musical tone generating apparatus according to the embodiment of the present invention. The horizontal axis indicates the time from the key-on information input, and the vertical axis indicates the DCO (musical sound generation channel) number. Simultaneously, three keys (keyA, keyB, keyC) are depressed, and a piano sound PN is generated in stereo (L / R) for each key. In this case, one tone color series is used. Here, it is assumed that a processing time of 1 ms is required for one DCO.

In FIGS. 21 and 22, the tone generation process is performed sequentially for each DCO. However, in this embodiment, a plurality of musical tones are simultaneously emitted after a predetermined time (6 ms) has elapsed from the input of the key-on information. Specifically, when key-on information of a plurality of musical tones is input, tone generator parameters of a plurality of musical tones are generated, sequentially assigned to a plurality of DCOs (sound sources) and set, and an operation stop flag of each DCO is set. This causes each DCO to wait for the start of sound generation. After a lapse of a predetermined time (6 ms) from the input of the key-on information, the calculation stop flags of all the DCOs are cleared, thereby instructing all the DCOs of the sound sources for which the sound source parameters are set to start sounding.

The first DCO is assigned a left (L) piano tone PN having a pitch A, and the second DCO is assigned a right (R) piano tone PN having a pitch A. . Third
Is assigned a left (L) piano tone PN with a pitch B, and the fourth DCO is assigned a right (R) piano tone PN with a pitch B. The left (L) piano sound PN having a pitch of C is assigned to the fifth DCO, and the sixth DCO is assigned to the sixth DCO.
Is assigned to the right (R) piano sound PN whose pitch is C.

As soon as the key-on information is input, the tone of each key-on information is assigned to the six DCOs, the parameters of each tone are generated, and each DCO of the sound source is generated.
And the operation stop flag of each DCO is set. Thereafter, when a predetermined time (6 ms) has elapsed from the input of the key-on information, the sound generation of the six DCOs is started by clearing the calculation stop flag of each DCO of the sound source. Assuming that a processing time of 1 ms per one DCO, processing of 6 DCOs takes 6 ms. The above-mentioned six DCO musical tones can be generated simultaneously at the time of 6 ms.

According to the present embodiment, sound generation is not started immediately after sound source parameters required for sound generation are set in the sound source, but a predetermined time (6 m) from input of key-on information.
Wait until s) elapses, during which time the sound source parameter calculation is completed, and when the sound source parameters for a plurality of musical tones are set in the sound source, the sound source is instructed to start sounding. it can. The player presses the three keys at the same time, and can simultaneously produce all sounds as intended by the player. This is not limited to the case where three keys are pressed at the same time, but the same applies when MIDI data and automatic performance data of three keys are input.

FIG. 2 is a time chart showing a second example of the tone generation processing of the electronic musical tone generator according to the present embodiment. The horizontal axis shows the time from the key-on information input, and the vertical axis shows the DCO number. At the same time, three keys (keyA, keyB, key
C) shows a case where a key is pressed and a piano sound PN and a strings sound ST are generated in stereo (L / R) for each key. In this case, two tone colors are used. Here, it is assumed that a processing time of 1 ms is required for one DCO.

The first DCO is assigned a left (L) piano tone PN having a pitch A, and the second DCO is assigned a right (R) piano tone PN having a pitch A. . Third
Is assigned a left (L) piano tone PN with a pitch B, and the fourth DCO is assigned a right (R) piano tone PN with a pitch B. The left (L) piano sound PN having a pitch of C is assigned to the fifth DCO, and the sixth DCO is assigned to the sixth DCO.
Is assigned to the right (R) piano sound PN whose pitch is C.

As soon as the key-on information is input, the musical tones of the key-on information are assigned to the first to sixth six DCOs, and the sound source parameters of the musical tones are generated. At the same time, the operation stop flag of each DCO is set. After that, when a predetermined time (6 ms) has elapsed from the input of the key-on information, the calculation stop flag of each DCO of the sound source is cleared, whereby
The sounding of the DCOs is started. Since the processing of the six DCOs takes 6 ms, only the six tones that can be emitted after a predetermined time (6 ms) have elapsed are first simultaneously emitted.

Next, the left (L) strings sound ST of pitch A is assigned to the seventh DCO, and the eighth DCO is assigned with the eighth DCO.
A string sound ST whose pitch is A to the right (R) is assigned to CO. At this time, sound source parameters of each musical tone are generated and set in each DCO, and a calculation stop flag of each DCO is set. Thereafter, after a lapse of 8 ms from the input of the key-on information, the operation stop flags of the seventh and eighth DCOs are cleared, whereby the seventh and eighth DCOs are cleared.
Produce the musical tone of CO. The tones of the seventh and eighth DCOs may be sequentially emitted at intervals of 1 ms, but since these tones are a set of stereo L / R, it is preferable that the tones are simultaneously emitted as in the present embodiment.

Next, a left (L) string sound ST with a pitch B is assigned to the ninth DCO, and a right (R) string sound ST with a pitch B is assigned to the tenth DCO.
Is assigned. At this time, sound source parameters of each musical tone are generated and set in each DCO, and a calculation stop flag of each DCO is set. Then, after a lapse of 10 ms from the input of the key-on information, the ninth and tenth DCOs
By clearing the calculation stop flag, the tone of the ninth and tenth DCOs is generated.

Next, a left (L) string sound ST with a pitch C is assigned to the eleventh DCO, and a right (R) string sound ST with a pitch C is assigned to the twelfth DCO.
Is assigned. At this time, sound source parameters of each musical tone are generated and set in each DCO, and a calculation stop flag of each DCO is set. Then, after a lapse of 12 ms from the input of the key-on information, the eleventh and twelfth DC
By clearing the O operation stop flag, the tones of the eleventh and twelfth DCOs are generated.

Here, it is conceivable to simultaneously generate the above-mentioned 12 DCO musical tones. That is, since the processing of the 12 DCOs takes 12 ms, the musical tones of the 12 DCOs can be generated simultaneously after a lapse of 12 ms from the input of the key-on information. However, in this case, since the sound is finally sounded 12 ms after the player performs the key pressing operation, it is necessary to consider whether or not the response time of the key pressing operation is within the permissible range of the player. . From this point of view, as in the present embodiment, it is preferable that, after a lapse of a predetermined time (6 ms) from the input of the key-on information, a predetermined number of musical tones (six sounds) be generated first, and then the remaining musical tones be generated. Even in this case, the twelfth DCO
Since it takes only 6 ms to generate CO, it can be said that the requirement for simultaneous sound generation is sufficiently satisfied as compared with the case of FIG. 21 which takes 10 ms.

In consideration of the response of the key pressing operation, the predetermined time is preferably 2 ms to 20 ms, and 3 ms to 20 ms.
10 ms is more preferred, and 6 ms is particularly preferred. The predetermined number of musical tones is determined by the processing capability of the CPU that performs the above-described processing. In the present embodiment, the processing time of 1 ms per DCO is assumed by the CPU, so that the predetermined time is 6 ms. Sometimes the predetermined number of musical tones is six. Therefore, the above-mentioned predetermined number of musical tones is not limited to six, but becomes the number of DCOs that can be processed within the above-mentioned predetermined time.

It is also possible to generate six piano sounds PN at a time of 6 ms from the input of the key-on information and to generate the remaining six strings sounds ST at a time of 12 ms. However, in this case, since the interval between the sounding of the six piano sounds PN and the sounding of the six strings sounds ST is a relatively long time of 6 ms, it is performed so that a total of 12 musical tones are not simultaneously sounded. May be heard by others. Therefore, as in the present embodiment, six piano sounds PN
, It is preferable that the remaining six string sounds ST be sequentially emitted in stereo sound units.

In this embodiment, a group of the piano sounds PN is generated first, and then a group of the strings sounds ST is generated. Conversely, the strings sound ST first
, And then the piano sound PN. However, in general, the attack speed (rising speed) of the musical tone envelope of the piano sound PN is higher than that of the strings sound ST, and thus the attack speed is higher.
It is preferable that the piano sounds PN be simultaneously generated first, and the strings sounds ST having the slower attack speed be sequentially generated later. That is, if a plurality of musical tones are pronounced with a slight shift in a musical tone having a fast attack speed, the listener is likely to sensitively sense a temporal shift in the musical tone. On the other hand, in the case of a musical tone having a slow attack speed, even if a plurality of musical tones are slightly shifted, the listener hardly feels a shift in the musical tone. Therefore, it is preferable that a musical tone with a faster attack speed be given a higher priority and sounded first.

Similarly, since a tone with a higher volume is more sensitive to a time lag of the tone, it is preferable that a tone with a higher volume has a higher priority and is sounded first.
In addition, priorities can be assigned according to key-on velocity and loudness used for controlling the attack speed and the volume. In this case, it is preferable that a tone having a higher velocity or loudness be given a higher priority and sounded first.

Further, priorities may be assigned according to pitches. Sounds in a frequency band around 1 kHz are heard with the highest sensitivity in human hearing. Therefore, it is possible to raise the priority of a musical tone having a pitch around 1 kHz and generate the tone first. For example, in the case of a 61-key keyboard instrument, a key having a pitch of 1 kHz belongs to a high pitch range. Therefore, the processing can be simplified, and the higher the tone, the higher the priority, and the tone can be generated first.

The parameters such as the above-mentioned attack speed and the like have standard values determined according to the timbre. Therefore, priorities may be assigned according to the timbre.

FIG. 3 is a time chart showing a third example of the tone generation process of the electronic musical tone generator according to the present embodiment. The horizontal axis shows the time from the key-on information input, and the vertical axis shows the DCO number. 4 keys at the same time (keyA, keyB, key
C, key D) is depressed, and a piano sound PN is generated in stereo (L / R) for each key. in this case,
One tone sequence will be used. Here, one DC
It is assumed that a processing time of 1 ms is required for O.

The first DCO is assigned a left (L) piano tone PN having a pitch A, and the second DCO is assigned a right (R) piano tone PN having a pitch A. . Third
Is assigned a left (L) piano tone PN with a pitch B, and the fourth DCO is assigned a right (R) piano tone PN with a pitch B. The left (L) piano sound PN having a pitch of C is assigned to the fifth DCO, and the sixth DCO is assigned to the sixth DCO.
Is assigned to the right (R) piano sound PN whose pitch is C.

As soon as the key-on information is input, the tone of each piece of key-on information is assigned to the first to sixth six DCOs, and tone generator parameters of each tone are generated. At the same time, the operation stop flag of each DCO is set. After that, when a predetermined time (6 ms) has elapsed from the input of the key-on information, the calculation stop flag of each DCO of the sound source is cleared, whereby
The sounding of the DCOs is started. Since it takes 6 ms to process the six DCOs, only six tones that can be emitted after a predetermined time (6 ms) have been emitted at the same time.

Next, a left (L) piano sound PN with a pitch of D is assigned to the seventh DCO, and a right (R) piano sound PN with a pitch of D is assigned to the eighth DCO. Is assigned. At this time, sound source parameters of each musical tone are generated and each D
CO is set, and the operation stop flag of each DCO is set. After that, 8m from input of key-on information
After the elapse of s, the tone stop of the seventh and eighth DCO is generated by clearing the operation stop flag of the seventh and eighth DCO. The tones of the seventh and eighth DCOs may be sequentially generated at intervals of 1 ms.
Since it is a set of / R, it is preferable to sound simultaneously as in this embodiment.

FIG. 4 is a time chart showing a fourth example of the tone generation process of the electronic musical tone generator according to the present embodiment. The horizontal axis shows the time from the key-on information input, and the vertical axis shows the DCO number. 6 keys (keyA, keyB, key)
C, keyD, keyE, and keyF) are depressed, and a divia sound TA and a string sound ST are generated in monaural for each key. In this case, two tone colors are used. Here, 1 ms for one DCO
Processing time.

The first DCO is assigned a tibia sound TA of pitch A, the second DCO is assigned a tibia sound TA of pitch B, and the third DCO is assigned a pitch sound TA. Is assigned a tibia tone TA of C, a fourth DCO is assigned a tibia tone TA of a pitch D, a fifth DCO is assigned a tibia tone TA of a pitch E, and a sixth DCO is assigned a tibia tone TA of a pitch E. Is assigned a tibia sound TA having a pitch of F.

As soon as the key-on information is input, the musical tones of the key-on information are sequentially assigned to the first to sixth six DCOs, and sound source parameters of the musical tones are generated. At the same time, the operation stop flag of each DCO is set. After that, when a predetermined time (6 ms) has elapsed from the input of the key-on information, the calculation stop flag of each DCO of the sound source is cleared, whereby
The sounding of the DCOs is started. Since it takes 6 ms to process the six DCOs, only six tones that can be emitted after a predetermined time (6 ms) have been emitted at the same time.

Next, similarly to the above, the seventh DCO includes:
A string sound ST having a pitch A is assigned, and after a lapse of 7 ms from the key-on information input, a tone of the seventh DCO is generated. Similarly, after a lapse of 8 ms from the key-on information input, the eighth DCO musical tone (key B string sound S
T), and after 9 ms from the input of the key-on information, the ninth DCO musical tone (string sound ST of keyC)
, And 10 ms after the key-on information input, the tenth DCO musical tone (keyD string sound ST)
, And 11 ms after the key-on information input, the eleventh DCO tone (key E string sound ST)
And the twelfth DCO musical tone (keyF string sound ST) 12 ms after the key-on information input
Is pronounced.

In this case, since the attack speed of the tibia sound TA is faster than that of the strings sound ST, the tibia sound T
A has a higher priority. In other words, the tibia sound TA with the fast attack speed is assigned to the first to sixth DCOs and sounded first, and thereafter the string sound ST with the slow attack speed is assigned to the seventh to twelfth DCOs and sounded. As described above, a tone with a fast attack speed has a higher priority for a tone with a faster attack speed, because even if a plurality of tones are pronounced with a slight shift, the listener perceives a time shift of the tone. To be pronounced.

FIGS. 5A and 5B are diagrams showing the format of the automatic performance data used in this embodiment.

FIG. 5A shows a sequence message (MIDI)
Channel message). The bit number is shown in the horizontal direction and the byte number is shown in the vertical direction. The 0th byte indicates a step type that takes a value from 00H to EFH in hexadecimal notation. The step time indicates the time from the immediately preceding bar (bar) message. The first byte is
Indicates status and sequence (MIDI channel). The status is based on MIDI, and indicates, for example, key-off, key-on, control change, program change, and the like. The second byte indicates a first data byte conforming to MIDI, for example, a key number. The third byte is a second data byte conforming to MIDI, and indicates, for example, velocity.

FIG. 5B shows the format of a system message. All system messages are 1 byte, and can be distinguished from the above sequence message in that the most significant bit becomes 1. Bar (bar) message SY
1 is represented by, for example, F8H. Automatic performance (SEQ)
The start message SY2 is represented by, for example, FAH. The automatic performance (SEQ) stop message SY3
For example, it is represented by FCH.

FIG. 6 shows automatic performance data (SEQ data) when the musical tone shown in FIG. 1 is generated. The sequence of numbers in the vertical direction indicates byte numbers. The 0th byte is an automatic performance start message SY2 (FIG. 5B), and the 1st byte is a bar message SY1. 2nd to 5th
Byte sets the tone of the piano sound. In the sixth to ninth bytes, when the step time is 3, a key-on of a musical tone having a pitch A is instructed. In the tenth to thirteenth bytes, the key-on of the musical tone with the pitch B is instructed when the step time is 3. In the 14th to 17th bytes, the key-on of the musical tone having the pitch C is instructed when the step time is 3. Thereafter, a key-off instruction corresponding to each key-on is issued, and finally, an automatic performance stop message SY3 (FIG. 5).
(B)) is located. This automatic performance data is
Instructs key-on of stereo sound of two pitches.

FIG. 7 shows automatic performance data when the tone shown in FIG. 2 is generated. The sequence of numbers in the vertical direction indicates byte numbers. The 0th byte is an automatic performance start message SY2 (FIG. 5B), and the 1st byte is a bar message SY1. In the second to fifth bytes, the timbre of the piano sound is set to the number 0 in the timbre series. In the sixth to ninth bytes, the timbre of the strings sound is set to the first in the timbre series.

In the tenth to thirteenth bytes, when the step time is 3, the pitch A is set to the 0th (piano sound) of the timbre series.
To key on the music. In the fourteenth to seventeenth bytes, when the step time is 3, the key-on of the tone having the pitch A is instructed to the first (strings sound) of the tone color series.

In the eighteenth to twenty-first bytes, when the step time is 3, the pitch is B at the 0th (piano sound) of the timbre series.
To key on the music. In the 22nd to 25th bytes, when the step time is 3, the key-on of the tone having the pitch B is instructed to the first (strings sound) of the tone color series.

In the 22nd to 25th bytes, when the step time is 3, the pitch is C at the 0th (piano sound) of the timbre series.
To key on the music. In the 26th to 29th bytes, when the step time is 3, the key-on of the tone having the pitch C is instructed to the first (strings sound) of the tone color series.
Thereafter, a key-off instruction corresponding to each key-on is issued, and finally, an automatic performance stop message SY3 (FIG. 5).
(B)) is located.

FIG. 8 is a block diagram showing the configuration of an electronic keyboard instrument (electronic musical sound generator) according to this embodiment. The bus 15 includes a timer 1, a CPU 2, a flash memory 3,
RAM 4, MIDI input / output interface 5, panel (operator / display) 6, key scan circuit 8, sound source LS
I10 and DSP12 are connected.

The timer 1 counts time information, and supplies a first interrupt signal INT1 to the CPU 2 when a predetermined time (6 ms) has been measured from the input of key-on information, and a time interval of (quarter note time) / 24 And the second interrupt signal I
NT2 is supplied to CPU2. The CPU 2 executes a computer program in the flash memory 3 corresponding to the flowcharts shown in FIGS.
The above simultaneous sound generation processing is realized.

The flash memory 3 stores a program processed by the CPU 2, a program processed by the DSP 12, automatic performance data (FIGS. 6 and 7), and timbre data. R
AM4 is a panel event buffer corresponding to the panel 6, a key event buffer corresponding to the keyboard 7, a panel 6
Has an analog event buffer corresponding to the analog operator in the inside.

The key scan circuit 8 performs a key scan on the keyboard 7 to detect key-on information, key-off information, velocity, pitch, and the like. The tone generator LSI 10 has a common operation start flag (CDCF: Cl) with the assignment memory 10a.
ear Disable Calculate Fla
g) having 10b. When the key-on information is input, the CPU 2 allocates the DCO, and sets the tone generator parameters corresponding to the key-on in the assignment memory 10a. The common operation start flag 10b is a flag for starting sound generation at the same time after a lapse of a predetermined time (6 ms) from input of key-on information.

The waveform memory 9 stores a PCM waveform of a musical tone. The sound source LSI 10 generates a musical tone by reading a waveform from the waveform memory 9. DSP12 is a CPU
2. Input the effect parameters from
An effect is applied to the musical tone output from 0 and output to the D / A converter 13. At this time, the DSP 12 gives an effect of delaying the musical sound using the delay memory 11.

The D / A converter 13 converts a musical tone input from the DSP 12 from a digital format to an analog format, and outputs it to an amplifier and a speaker 14. A tone is generated from the amplifier and the speaker 14.

FIG. 9 is a diagram showing the contents of the assignment memory 10a in the sound source LSI 10 shown in FIG. The numbers arranged in the vertical direction are addresses. At addresses 000H to 007H, a parameter having a DCO number of 0 is stored.
A parameter having a DCO number of 1 is stored in addresses 8H to 00FH, and a DCO number is stored in addresses 1F8H to 1FFH.
The parameter with the number 63 is stored. Thus, D
Since the addresses are sequentially stored in the order of the CO number, simply incrementing the address enables
All DCO parameters can be accessed.

The parameters of each CDO are, for example, waveform start address, F number (frequency number), filter parameter, envelope parameter (attack level and attack speed, etc.), loudness (volume), panning (left and right localization), F number Accumulator (FAC
C), envelope accumulator (EACC), DC
A flag USE indicating whether O is in use, a series, an interpolation flag IPF, and a calculation stop flag (DCF: Disable
Calculate Flag).

FIG. 10 is a block diagram showing the configuration of the sound source LSI 10 of FIG. F number interpolation circuit 21, filter parameter interpolation circuit 22, loudness interpolation circuit 23,
When the interpolation flag IPF is 1, the panning interpolation circuit 24 interpolates the F number, the filter parameter, the loudness, and the panning in the assignment memory 10a, and when the interpolation flag is 0, outputs the data without interpolation. These interpolations are performed to alleviate a sudden change from the sound that was previously generated.

The F-number accumulating circuit 25 accumulates the F-number input from the F-number interpolating circuit 21 and outputs an integer part of the accumulated value to the adder 33,
Store the F number accumulator FACC in 0a.
The adder 33 adds the accumulated integer part of the F number and the waveform start address in the assignment memory 10a, and outputs the result to the waveform memory 9.

The waveform memory 9 reads out the internal waveform based on the above-mentioned added value and outputs it to the sample interpolation circuit 27. The sample interpolation circuit 27 interpolates the waveform read from the waveform memory 9 based on the decimal part of the accumulated value input from the F number accumulation circuit 25 and outputs the interpolated signal to the filter 28. The filter 28 has a function of a sample interpolation circuit 27 according to a filter parameter input from the filter parameter interpolation circuit 22.
Filters the output waveform.

The envelope generation circuit 26 generates a musical tone envelope based on the envelope parameters stored in the assignment memory 10a, and outputs it to the multiplier 29 and the envelope accumulator EACC of the assignment memory 10a. The multiplier 29 multiplies the waveform output from the filter 28 by the envelope output from the envelope generation circuit 26, and outputs the product.

The multiplier 30 multiplies the waveform output from the multiplier 29 by the loudness value output from the loudness interpolation circuit 23 and outputs the product. The multiplier 31 multiplies the waveform output from the multiplier 30 by the panning value output from the panning interpolation circuit 24 and outputs the result.

The sequence distribution and sequence accumulation circuit 32 distributes, based on the sequences stored in the assignment memory 10a, the time-divided waveforms of the tone generation channels output from the multiplier 31 to the respective sequences. And accumulates for each series, and outputs, for example, four series of waveforms to the DSP 12 (FIG. 8). For example, the number of tone generation channels is 64, and the sequence is 4.

The operation stop flag (DCF) in the assignment memory 10a is provided for each tone generating channel. When the key-on information is input, a parameter is generated and a tone generation channel is allocated. Then, the generated parameter is set in the tone generation channel in the assignment memory 10a, and a calculation stop flag (DCF) is set.
To stop the calculation and wait for the sound to start.
After a lapse of a predetermined time (6 ms) from the input of the key-on information, the computation stop flag (DCF) is cleared and sound generation is started.

Specifically, when the operation stop flag (DCF) is set, the value of the operation stop flag (DCF) is supplied to the envelope generation circuit 26 and the F number accumulation circuit 25. Then, the envelope generation circuit 26 sets the envelope value to 0 and outputs the same, and the F number accumulation circuit 25 also sets the F number to 0 and outputs the same. By setting the envelope value to 0, no musical tone is generated, and by setting the F number to 0, the start of sound generation is awaited. Next, the operation stop flag (D
When (CF) is cleared, the envelope generation circuit 26 and the F number accumulation circuit 25 start generating and outputting the normal envelope and F number, respectively. That is, the start of generation of a musical tone is instructed.

When the common operation start flag (CDCF) 10b is set, the operation stop flags (DCF) of all tone generation channels in the assignment memory 10a are cleared, and generation of all tones is started.

FIG. 11 is a flowchart of a main routine executed by the CPU. When the power is turned on, initialization processing such as setting to a register and permitting an interrupt is performed in step SA1.

At step SA2, the panel 6 (FIG. 8) is scanned, an event generated by the player's operation of the panel is detected, and status (operation type) and sequence data are added and stored in the panel event buffer.

In step SA3, of the events generated by the above-described panel scan, an event by an analog operator, such as a volume, which is processed in an analog manner,
Status and sequence data are added and stored in the analog event buffer.

At step SA4, the keyboard 7 (FIG. 8) is key-scanned to detect an event generated by a player's key operation, and to add a status (key-on / key-off), velocity and sequence data to the key event. Store in buffer.

At Step SA5, it is determined whether or not an automatic performance is instructed. When the automatic performance is instructed, the reproduction processing of the automatic performance is performed in step SA6, and the process proceeds to step SA7. When the automatic performance has not been instructed, the process directly proceeds to step SA7. The details of step SA6 will be described later with reference to the flowcharts of FIGS.

At step SA7, panel event processing is performed. Details of this processing will be described later with reference to the flowchart of FIG.

At step SA8, key event processing is performed. Details of this processing will be described later with reference to FIGS.
This will be described with reference to the flowchart of FIG.

At Step SA9, analog event processing is performed. Details of this processing will be described later with reference to the flowchart of FIG.

At step SA10, the upper one of the envelope accumulators (EACC) for the DCO (tone generating channel) whose key is OFF and whose flag USE is 1 is set.
It is checked whether the 0 bit is 0 or not. If it is 0, the flag USE of the DCO is set to 0. That is, when the envelope value is close to 0, a new key-on can be allocated to the DCO by setting the flag USE to 0.

In step SA11, other processing is performed, and the flow returns to step SA2 to repeat the above processing.

FIG. 12 is a flowchart showing the details of the automatic performance reproduction process shown in step SA6 of FIG.

In step SB1, the flag waitBA
Check whether R is 0 or not. Flag waitBAR
Means that when a bar message is read,
It is set to 1 to indicate that the bar is being processed, and then cleared to 0 by a timer interrupt shown in FIG. When the flag waitBAR is 1, the process ends without doing anything. When the flag waitBAR is 0, the process proceeds to Step SB2.

At step SB2, it is checked whether the next message to be read is a series message (FIG. 5A) or a system message (FIG. 5B). When the message is a series message, the process proceeds to step SB3, and when the message is a system message, the process proceeds to step SB13 in FIG.

In step SB3, only the step time in the message indicated by the automatic performance read pointer is read.

At Step SB4, it is checked whether or not the register Beat is equal to or longer than the step time read as described above. The register Beat is the current number of beats counted by the timer interrupt in FIG. When the value of the register Beat is less than the step time, the processing ends because the processing timing has not yet been reached. If the value of the register Beat is equal to or longer than the step time, it means that the processing timing has been reached, and the process proceeds to step SB5.

At step SB5, the message indicated by the automatic performance read pointer is read, and at step SB6, the automatic performance read pointer is advanced to the position of the next message.

At step SB7, it is checked whether the read message is a program change or a control change. If it is a program change or a control change, the process proceeds to step SB8. If it is not a program change or a control change, the process proceeds to step SB9.

At step SB8, the step time is deleted from the read message and stored in the panel event buffer, and the process returns to step SB1.

In step SB9, it is checked whether the read message is key-on or key-off. If the key is on or off, step S
Proceed to B10. When not key-on or key-off,
Proceed to step SB11 in FIG.

At step SB10, the step time is deleted from the read message and stored in the key event buffer, and the process returns to step SB1.

In step SB11 in FIG. 13, it is checked whether the read message is a key pressure, a channel pressure or a pitch bend. When the above message is satisfied, the process proceeds to step SB12,
If not, the process proceeds to Step SB22.

At step SB12, the step time is deleted from the read message and stored in the analog event buffer, and the process returns to step SB1 of FIG.

At step SB22, other series messages are processed, and the process returns to step SB1 of FIG.

Step SB13 is a process for a system message. It is checked whether or not the message indicated by the automatic performance read pointer is a bar message. When the message is a bar message, the process proceeds to step SB14. When the message is not a bar message, the process proceeds to step SB16.
Proceed to.

At step SB14, the flag waitBAR is set to 1 according to the bar message, and at step SB1
In step 5, the automatic performance read pointer is advanced to the next message position, and the process is terminated.

In step SB16, it is checked whether or not the message indicated by the automatic performance read pointer is an automatic performance start message. If the message is an automatic performance start message, the process proceeds to step SB17. If the message is not an automatic performance start message, the process proceeds to step SB19.
Proceed to.

At Step SB17, the flag SEQ_P
LY is set to 1, the automatic performance read pointer is advanced to the next message position in step SB18, and the process returns to step SB1 in FIG.

At step SB19, it is checked whether or not the message indicated by the automatic performance read pointer is an automatic performance stop message. If the message is not an automatic performance stop message, the process proceeds to step SB20. If the message is not an automatic performance stop message, the process proceeds to step SB23.
Proceed to.

At step SB20, the flag SEQ_P
LY is set to 0, and in step SB21, all the DCOs that are sounding as the key-on of the automatic performance data replace the envelope with the release mode (silence processing mode). After that, the process ends.

At step SB23, other system messages are processed, and the process ends.

FIG. 14 is a flowchart showing details of the panel event process shown in step SA7 of FIG.

At step SC1, it is checked whether or not the panel event buffer is empty. If it is empty, the process ends. If it is not empty, the process proceeds to step SC2.

In step SC2, one event is read from the panel event buffer, and in step SC3,
Check whether the read event is a program change. Step SC when program change
The process proceeds to step SC5, and if not a program change, the process proceeds to step SC5.

In step SC4, a preparation for switching the timbre of the corresponding sequence is made. That is, the tone currently being produced in the series continues to be produced in the same tone color, and the tone of the corresponding series is switched and produced after the next occurrence of a new key-on. Thereafter, the process returns to step SC1.

At step SC5, it is checked whether or not the read event is a control change. If it is a control change, the process proceeds to step SC6. If it is not a control change, the process proceeds to step SC7.

At step SC6, the sound quality, pitch, panning, reverb (reverb relates to the DSP program and the coefficient) of the corresponding sequence are controlled based on the control change information. Thereafter, the process returns to step SC1.

At step SC7, other events are processed, and the process returns to step SC1. Thereafter, the above processing is repeated until the panel event buffer becomes empty.

FIG. 15 is a flowchart showing details of the analog event processing shown in step SA9 of FIG. Analog events are, for example, volume (volume control), aftertouch (volume and / or timbre control), and vendor (pitch and / or timbre control).

In step SE1, it is checked whether the analog event buffer is empty. If it is empty, the process ends. If it is not empty, the process proceeds to step SE2.

In step SE2, one event is read from the analog event buffer. In step SE3, the analog event information is converted into parameters for setting in the assignment memory.

At step SE4, the interpolation flag IPF is set to 1 for each DCO in the assignment memory to which the corresponding sequence is assigned. Regarding analog events, it is preferable to reduce sudden changes in parameters by interpolating the parameters.

At step SE5, the above-mentioned new parameters are set for each DCO of the assignment memory to which the corresponding sequence is assigned. Then, step SE1
And the above processing is repeated. When the analog event buffer becomes empty, the processing is terminated.

FIG. 16 is a flowchart showing details of the key event processing shown in step SA8 of FIG.

At step SD1, it is checked whether or not the key event buffer is empty. If not empty, proceed to step SD2.

At step SD2, one event is read from the key event buffer. At step SD3, it is checked whether the read event is a key-off event. If it is not a key-off event, it is a key-on event, and the flow advances to step SD8.

At step SD8, it is checked whether or not TIMER1 is counting. TIMER1 is a timer that generates an interrupt signal 6 ms after the input of the key-on information, and is processed in FIGS. 19A and 19B. When counting is not in progress, TIMER1 is kept for a predetermined time (6 m
s) is set, TIMER1 is started to count down, and the flow advances to step SD10. If counting is in progress, step SD9 is bypassed and step SD1 is skipped.
Go to 0. That is, when a plurality of key-on events are input, measurement is performed for a predetermined time (6 ms) from the time when the first key-on event is input.

In step SD10, the attack speed (speed) of the tone set in the corresponding series is added and stored in the allocation priority buffer. When a tone uses 2 DCOs per key-on, the tone is decomposed and stored in DCO units. Thereafter, the process returns to step SD1.

If it is determined in step SD3 that the event is a key-off event, the flow advances to step SD4.

At step SD4, it is checked whether or not the corresponding DCO is already producing sound. When the sound is being produced, step S is performed to perform the normal key-off process.
Proceed to D7.

At step SD7, the envelope parameter of the corresponding DCO is set as a key-off release parameter, and a mute process is performed. Then, step S
Return to D1.

If it is determined in step SD4 that the sound is not being generated, the flow advances to step SD5 to check whether or not the sound is allocated but before the sound is started. If it has been assigned, the corresponding DCO
Clear the flag USE and the computation stop flag DCF to 0, and return to step SD1. If not, step SD6 is bypassed and the process returns to step SD1.

If it is determined in step SD1 that the key event buffer is empty, the flow advances to step SD11.

At step SD11, it is checked whether or not the flag OVER is "1". The flag OVER becomes 1 when the number of key-ons exceeding the maximum number of DCOs (6 DCOs) capable of simultaneous sound generation is accumulated in the allocation priority buffer.
When the flag OVER is 0, step SD in FIG.
13 and when the flag OVER is 1, the process proceeds to step SD17 in FIG.

At step SD13 in FIG. 17, it is checked whether or not the allocation priority buffer is empty. If it is not empty, the process proceeds to step SD14, and if it is empty, the process ends.

In step SD14, the DCO assignment of the sound source LSI and the calculation of various parameters are performed with priority given to the key-on information of the tone having the fastest attack speed of the envelope, and the result is set in the corresponding assignment memory. If the assignment memory is not empty, a specific DCO is truncated (key-off and high-speed release) according to a predetermined priority. Next, the number of assigned DCOs is stored in the register waitDCO.
To store. Next, the key-on event for which the assignment has been completed is deleted from the assignment priority buffer. Next, all the assigned DCOs set the flag USE to 1, the interpolation flag IPF to 0, and the calculation stop flag DCF to 1. By setting the computation stop flag DCF to 1, the apparatus enters a standby state without starting computation for sound generation. Thereafter, the process proceeds to Step SD15.

In step SD15, the DC in the standby state for sound generation is determined.
It is checked whether or not the register waitDCO indicating the number of O is 6 or less. When the value of the register waitDCO is equal to or smaller than 6, the process returns to step SD13 to repeat the above processing. When the value of the register waitDCO exceeds 6, the process proceeds to Step SD16. In step SD16, the flag OVER is set to 1, and the process is terminated.

The step SD17 in FIG.
When ER is set to 1 and a key-on number exceeding the number of DCOs that can be simultaneously sounded is assigned, this is a process for starting sounding of the currently assigned DDCO. It is determined whether or not the assignment priority buffer is empty. To check. If the allocation priority buffer is not empty, the process proceeds to step SD18.

At step SD18, an event of 1DCO is read from the allocation priority buffer. Step SD19
In this example, the DCO assignment of the sound source LSI and the calculation of various parameters are performed with priority given to the key-on information of the timbre having the fast attack speed of the envelope, and the result is set in the corresponding assignment memory. If the assignment memory is not empty, a specific DCO is truncated (key-off and high-speed release processing) according to a predetermined priority. Next, the key-on event for which the assignment has been completed is deleted from the assignment priority buffer. Next, all the assigned DCOs set the flag USE to 1, set the interpolation flag IPF to 0, and set the operation stop flag DCF
To 0. By setting the operation stop flag DCF to 0, the start of sounding operation is instructed, and sounding is started immediately.
In the case of stereo sound, left sound (L sound) and right sound (R sound)
Are successively read, and at the same time, an instruction to start sounding is issued.

At step SD20, it is checked whether or not the allocation priority buffer is empty. When the allocation priority buffer is not empty, the process returns to step SD18, and when the allocation priority buffer is empty, the process ends.

If the allocation priority buffer is empty in step SD17, the flow advances to step SD21. In step SD21, the flag OVER is set to 0, and the process ends.

FIG. 19A shows a first timer (TIM).
It is a flowchart which shows the 1st interruption processing example of ER1). The first timer starts counting down when 6 ms is set in step SD9 in FIG.
After ms, the interrupt signal is supplied to the CPU. The CPU
Upon receiving the interrupt signal, the following processing is performed.

First, in step SF1, the standby DC
0 is set in a register waitDCO indicating the number of Os.
Next, in step SF2, 1 is set to a common operation start flag CDCF of the sound source LSI. When the common operation start flag CDCF is set to 1, the sound source LSI
Is cleared to 0. Then, all the DCOs are released from the operation suspension, and the assigned DC
The start of O sound generation is instructed. That is, the sound generation start is instructed after a lapse of a predetermined time (6 ms) from the input of key-on information. When key-on information of a plurality of musical tones is input, the start of sound generation is instructed after a lapse of a predetermined time from the input of the first key-on information. And finally, the sound source LSI is
The common operation start flag CDCF is cleared to 0.

FIG. 19B shows a first timer (TIM)
It is a flowchart which shows the 2nd interruption processing example of ER1). In the above-described first interrupt processing example, the case where the sound source LSI clears the calculation stop flags DCF of all the DCOs to 0 when the common calculation start flag CDCF is set to 1 has been described. However, in the second interrupt processing example, The case where the CPU directly clears the operation stop flags DCF of all DCOs of the sound source LSI to 0 will be described. As in the first interrupt processing example, when the CPU receives the above-described interrupt signal, the CPU performs the following processing.

First, at step SG1, the standby DC
0 is set in a register waitDCO indicating the number of Os.
Next, in step SG2, interrupts are disabled (DI: Di
(sable interrupt). Next, in step SG2, the calculation stop flag DCF of each DCO of the sound source LSI is cleared to 0 in order. Next, in step SG4, an interrupt is enabled (EI: Enable In).
(Terrupt) state, and the process ends.

In the first interrupt processing example, a circuit is required for the tone generator LSI to clear the operation stop flags DCF of all DCOs in accordance with the common operation start flag CDCF.
The cost of the sound source LSI increases. If it is desired to reduce the cost of the sound source LSI, the CPU may directly clear the computation stop flag DCF as in the second interrupt processing example.

FIG. 20 shows a second timer (TIMER).
It is a flowchart which shows the interruption process example of 2). The second timer is a tempo timer for counting the beat register Beat used in step SB1 in FIG. 12, and supplies an interrupt signal to the CPU at intervals of, for example, 1/24 of a quarter note length. Upon receiving the interrupt signal, the CPU performs the following processing.

At step SH1, the beat register Be
Increment the value of at. Next, step SH2
Then, it is checked whether or not the value of the beat register Beat is 24 × 4 or more. That is, the interrupt interval is 4
In the case of 1 / 24th of the note length and 4/4 time signature,
When the value of the beat register Beat is 24 × 4, which means the length of one bar, the process ends when the value of the beat register Beat is less than 24 × 4, and proceeds to step SH3 when the value of the beat register Beat is 24 × 4 or more.

At step SH3, the bar register BAR
Is incremented. Next, the register waitBAR is set to 0 in step SH4, and the beat register Beat is set to 0 in step SH5. After that, the process ends.

According to this embodiment, the sound generation is not started immediately after the parameters required for sound generation are set in the sound source, but instead, the calculation stop flag DCF of each DCO is set to 1 and the predetermined time from the input of the key-on information is set. When the parameters for a plurality of musical tones are set in the tone generator, the calculation stop flag DCF of each DCO is set to 0.
By instructing the sound source to start sounding by clearing the tone, a plurality of musical tones can be sounded simultaneously or in a short time.

When a plurality of musical tones to be produced simultaneously are to be produced, a predetermined priority order is assigned according to the timbre, velocity, pitch, volume or attack speed of the musical tone envelope, so that the sound production process is performed simultaneously. The listener can be made to feel as if a plurality of tones to be played were produced simultaneously or in a short time.

The present invention can also be implemented by supplying a program code of software for realizing the functions of the above embodiment and operating it according to a program stored in a computer (CPU or MPU) of the electronic musical sound generating apparatus.

In this case, the program code itself of the software realizes the function of the above-described embodiment, and the program code itself and means for supplying the program code to the computer, for example, storing the program code The recorded recording medium constitutes the present invention. As a recording medium for storing such a program code, for example, a floppy (registered trademark) disk, hard disk, optical disk, magneto-optical disk, CD-RO
M, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.

The above embodiments are merely examples of implementation of the present invention, and the technical scope of the present invention should not be interpreted in a limited manner. It is. That is, the present invention does not depart from the spirit or the main features thereof,
It can be implemented in various forms. For example, the present invention is not limited to the PCM sound source, but may be applied to other sound sources such as a sound source board built in a personal computer, a soft sound source that generates a waveform by a computer program, and the like.

[0154]

As described above, according to the present invention, the sound generation is not started immediately after the parameters required for sound generation are set in the sound source, but rather, a predetermined time is elapsed from the input of the key-on information. When the parameters for a plurality of musical tones are set in the sound source, the sound source is instructed to start sounding, so that a plurality of musical sounds can be produced simultaneously or in a short time.

[Brief description of the drawings]

FIG. 1 is a first view of an electronic musical sound generator according to an embodiment of the present invention.
3 is a time chart showing an example of a tone generation process.

FIG. 2 is a time chart showing a second example of a tone generation process of the electronic musical sound generating device according to the embodiment.

FIG. 3 is a time chart illustrating a third example of a tone generation process of the electronic musical sound generating device according to the present embodiment.

FIG. 4 is a time chart illustrating a fourth example of a tone generation process of the electronic musical sound generating device according to the embodiment.

FIGS. 5A and 5B are diagrams showing examples of a format of automatic performance data. FIG.

6 is a diagram showing automatic performance data corresponding to the example of the sound generation process shown in FIG. 1;

FIG. 7 is a diagram showing automatic performance data corresponding to the sound generation example shown in FIG. 2;

FIG. 8 is a block diagram showing a configuration of an electronic musical sound generating device (electronic keyboard instrument) according to the present embodiment.

FIG. 9 is a diagram showing a configuration of an assignment memory of a sound source LSI.

FIG. 10 is a block diagram illustrating a configuration of a sound source LSI.

FIG. 11 is a flowchart showing processing of a main routine performed by a CPU.

FIG. 12 is a flowchart showing details of automatic performance reproduction processing.

FIG. 13 is a flowchart showing details of the automatic performance reproduction process following FIG. 12;

FIG. 14 is a flowchart illustrating details of panel event processing.

FIG. 15 is a flowchart illustrating details of analog event processing.

FIG. 16 is a flowchart showing details of a key event process.

FIG. 17 is a flowchart showing details of the key event process following FIG. 16;

FIG. 18 is a flowchart illustrating details of another key event process following FIG. 16;

FIGS. 19A and 19B are flowcharts showing interrupt processing of the first timer.

FIG. 20 is a flowchart showing interrupt processing of a second timer.

FIG. 21 is a time chart showing a first example of a tone generation process of the electronic musical sound generator according to the prior art.

FIG. 22 is a time chart showing a second example of a tone generation process of the electronic musical tone generator according to the prior art.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Timer 2 CPU 3 Flash memory 4 RAM 5 MIDI input / output interface 6 Panel (operator / display) 7 Keyboard 8 Key scan circuit 9 Waveform memory 10 Sound source LSI 10a Assignment memory 10b Common operation start flag 11 Delay memory 12 DSP 13 D / A converter 14 Amplifier and speaker 15 Bus 21 F number interpolation circuit 22 Filter parameter interpolation circuit 23 Loudness interpolation circuit 24 Panning interpolation circuit 25 F number accumulation circuit 26 Envelope generation circuit 27 Sample interpolation circuit 28 Filter 29, 30, 31 Multiplier 32 Sequence distribution and sequence accumulation circuit 33 Adder

Claims (10)

[Claims] An input unit for inputting key-on information; a timer for starting measurement of a predetermined time from the input of the key-on information when the key-on information is input; and a parameter required for sound generation when the key-on information is input. An electronic musical tone generating apparatus comprising: parameter setting means for setting a sound source; and sound generation start instructing means for instructing a sound source to which the parameter is set to start sound generation after a lapse of a predetermined time from the input of the key-on information. 2. The electronic musical sound generator according to claim 1, wherein said parameter setting means, upon inputting said key-on information, generates parameters necessary for sound generation based on said key-on information and sets the parameters in a sound source. 3. The electronic musical sound generator according to claim 1, wherein the predetermined time is 2 ms to 20 ms. 4. The parameter setting means assigns and sets the parameter to one of the tone generating channels of the sound source, and sets a calculation stop flag of the tone generating channel to wait for sound generation to start. 2. The electronic musical sound generator according to claim 1, wherein the sounding start instructing means instructs sounding start by clearing the calculation stop flag after the predetermined time has elapsed. 5. The apparatus according to claim 1, further comprising a common operation start flag for simultaneously instructing a plurality of predetermined tone generation channels to start sounding, wherein said parameter setting means sets a plurality of tone parameters to a plurality of tone generation channels. When the common operation start flag is set, the sound generation start instructing means clears all the operation stop flags of the plurality of predetermined tone generation channels, and starts sound generation of the predetermined plurality of tone generation channels. 5. The electronic musical sound generator according to claim 4, wherein the electronic musical sound generator is instructed. 6. The sound generation start instructing means, when key-on information of a plurality of music sounds is input to the input means, collectively starts sound generation of only a predetermined number of music sounds among the input plurality of music sounds. 2. The electronic musical tone generating apparatus according to claim 1, wherein the electronic musical tone generating apparatus is instructed to start sounding after that. 7. The sound generation start instructing means for sequentially instructing the start of sound generation for the remaining musical tones thereafter.
The electronic musical sound generator according to the above. 8. The electronic musical tone generator according to claim 6, wherein said parameter setting means and said tone generation start instructing means are realized by processing of a CPU, and said predetermined number of musical tones is determined by a processing capability of said CPU. 9. A step of: (a) setting parameters required for sound generation in a sound source based on key-on information when key-on information is input; and (b) a sound source in which the parameters are set after a lapse of a predetermined time from the input of key-on information. Instructing the user to start sounding. 10. A procedure for setting parameters necessary for sound generation in a sound source based on key-on information when (a) key-on information is input; and (b) a sound source in which the parameters are set after a lapse of a predetermined time from the input of key-on information. And a computer-readable recording medium storing a program for causing a computer to execute a procedure for instructing a computer to start sounding.