JP2646923B2 - Automatic performance device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic performance apparatus for performing a plurality of automatic performances, and more particularly, to processing of automatic performance data of each type in accordance with a resolution set for each type, thereby storing performance data. The present invention relates to an automatic performance device capable of reducing memory.

[0002]

In a conventional, previously stored in the memory in accordance with elapsed playing data for a plurality of types of automatic performance was paired <br/> respond to performance parts plurality fraction (the tone data) time, the plurality 2. Description of the Related Art There is known an automatic performance apparatus which performs automatic performance by reading out various types of performance data for automatic performance simultaneously or selectively and sequentially according to the passage of time to generate musical tones.

[0003]

In the above-mentioned conventional automatic performance device, in order to generate musical tones corresponding to various notes in each type of automatic performance, the most important of the sounding timing and the note length of each note is generated. Since the resolution was set according to the fine sounding timing and note length, and the performance data for all notes was expressed according to this resolution, a large capacity memory was required to store the performance data. was there. Note that the resolution indicates how accurately the tone generation timing and tone generation time of a musical tone can be reproduced. That is, for example, if the shortest note length among the notes in each type of automatic performance is a sixteenth note, the resolution is naturally set so that the sixteenth note can be expressed. Eighth note will be represented by "resolution = sixteenth note"
The performance data related to the eighth note has twice the storage amount as the performance data related to the sixteenth note. In the case of "resolution = sixteenth note", one storage location (address) of the memory corresponds to the sixteenth note, and the performance data relating to the sixteenth note is stored in one storage location. However, since the eighth note is twice as long as the sixteenth note, the performance data of the eighth note must be stored over two storage locations. Furthermore, performance data relating to quarter notes must be stored over four storage locations.

Therefore, in the conventional automatic performance device, all the performance data are processed based on one common resolution, so that the capacity of the performance data memory is increased. The same applies to the case where performance data is stored by an event method well known to those skilled in the art. The event-type performance data includes data representing a musical tone to be produced and interval time data representing a production timing (production interval time of each musical tone). In this case, the resolution depends on a time length that can be represented by the interval time. Decided. If this interval time is set corresponding to "resolution = sixteenth note", the interval time corresponding to the eighth note is twice as large as that of the sixteenth note, and the number of bits of the interval time data increases. As a result, the capacity of the performance data memory increases.

The present invention has been made in view of such circumstances, and in an automatic performance apparatus capable of performing a plurality of types of automatic performances, the circuit scale can be reduced by saving a memory for storing performance data. The purpose is to improve the realization and real-time processing.

[0006]

According to the first aspect of the present invention, there is provided an automatic performance apparatus for sequentially reading performance data stored in advance and performing an automatic performance in accordance with the read data. Read at the same time
Storage means for storing the performance data comprising a plurality of parts that are, the resolution data that determines the resolution of the sounding timing and sounding time of the tone corresponding to the previous SL performance data, sets the <br/> to each part Resolution setting means and each of the parts
Based on the resolution data, the performance data of each corresponding part characterized by comprising a read control means for reading in parallel at a cycle corresponding to the resolution. Claim 2
According to the invention described in (1), in an automatic performance device that sequentially reads performance data stored in advance and performs an automatic performance according to the read data, a plurality of data that are simultaneously read are stored .
Storage means for storing performance data composed of part-sounding timing and sounding time of the tone corresponding to the previous SL <br/> performance data
A resolution setting unit for the resolution data to determine the resolution is set to each part, the resolution de set for each part
A plurality of timekeeping means for performing timekeeping at a cycle corresponding to data,
Based on timing timing measured by each timing means
Te, characterized by comprising a read control means for reading the performance data of each corresponding part.

[0007]

According to the above arrangement, the resolution data for the performance data is set for each automatic performance part . Each path corresponds according resolution set for each Each Part
The timing of reading the performance data of the note is controlled. Therefore, the performance data of each part
Since it is possible to express each part with an optimum resolution in accordance with notes and the like in the automatic performance of the parts , it is possible to save the capacity of the performance data memory.

[0008]

An embodiment of the present invention will be described below with reference to the drawings. The time unit used in this embodiment is all milliseconds (hereinafter, referred to as msec).
It is a unit. A: Configuration of Embodiment FIG. 1 shows an embodiment of the present invention, and is a block diagram showing a configuration of an automatic performance device. In this figure, 1 is a keyboard. The keyboard 1 has a mechanism for detecting a pressed / released key and a speed of the pressed / released key for each key. Reference numeral 2 denotes a key event detection unit which scans each key of the keyboard 1 to detect an event of the key, and generates tone information for the key having the event each time an event is detected.
Output to The musical tone information includes a key code KC representing a note name (that is, a pitch), and a key-on signal K representing a key press event.
ON, a key-off signal KOFF indicating a key release event, and touch data indicating a key pressing or key releasing speed.

Reference numeral 3 denotes a switch group composed of various switches, and includes an automatic performance switch, a performance style selection switch, and the like. Reference numeral 4 denotes a switch event detection unit which generates switch event information, that is, a state change signal of each switch, and outputs the signal to the CPU 7. Reference numeral 8 denotes a timer which outputs an interrupt request signal to the CP every 1 msec.
It is sent to U7. Reference numeral 6 denotes a program memory which stores various control programs to be loaded into the CPU 7, various data used in these programs, and the like. Reference numeral 5 denotes a working memory, which is a memory for temporarily storing various data used for the operation of the CPU 7 and an operation result, and in which various registers described later are set.

The CPU 7 sends an operation control command to each section of the device based on the command loaded from the program memory 6. Also, each time the above-mentioned interrupt request signal of the timer 8 is generated, an interrupt control instruction routine is started, and interrupt control related to tone generation / silence timing control, tone generation time monitoring control, and the like is performed. Reference numeral 9 denotes a ROM in which data for automatic performance is stored. Automatic performance is this R
This is performed by referring to the performance data stored in the OM 9. Reference numeral 10 denotes a tone generator. The tone generator 10 generates the key-on signal KO described above.
A tone waveform is formed based on N, the key-off signal KOFF, the value of the key code KC, and the like, and is output to the sound system 9. The sound system 9 converts a digital waveform into an analog waveform and performs actual tone generation processing.

Next, the configuration of the ROM 9 will be described. First, as shown in FIG. 1, the ROM 9 has a style header 9a for storing a performance style, a code pattern area 9b for storing a chord progression, a pace pattern area 9c for storing a bass performance mode, and a rhythm pattern area 9d for storing a rhythm performance mode. It is composed of Next, these storage areas will be described.

First, the style header 9a will be described with reference to FIG. The mapping state of the style header 9a is as shown in the figure, and style headers corresponding to various automatic performance styles are stored in order. The style header 9a includes various data for specifying various automatic performance styles, such as beat BT, tempo TMP,
The tone color VOICE1 of the bass part, the tone color VOICE2 of the chord 1 part, the tone color VOICE3 of the chord 2 part,
Rhythm part resolution QT0, bass part resolution Q
T1, resolution QT2 of chord 1 part, resolution QT3 of chord 2 part, and pitch start point BGN are stored. Next, these data will be described.

First, the beat BT is used to designate a "number of beats", for example, a quarter beat, a triple beat, and the like. This data is used for "calculation of the number of measures" when reproducing a musical sound. Next, the tempo TMP is data for determining how many times a quarter note is to be pronounced equally per minute. The tempo TMP is data used for various timing controls such as calculation of a “division period for quarter notes” described later. Tones VOICE1-3
These are data for deciding what tone color the bass part, chord 1 part and chord 2 part are to be pronounced. The pitch start point BGN is data representing a pitch start point. The pitch start point BGN is data for determining the key code KC of the generated musical tone.

In the present embodiment, the values of the resolution data QT0 to QT3 are such that the sounding timing of the sound to be generated in the corresponding part and the resolution of the note length are an eighth note system,
It is determined by whether it is a triplet system, a sixteenth note system, or a sixt system. The resolution data QT0 to QT3 are set to “0” in the case of the eighth note system and the triplet system, and the resolution data QT0 are set in the case of the sixteenth note system and the sixt system. “1” is set as the value of QT3. In this case, in the resolution of the eighth note system and the triplet system, in one of six sounding units obtained by dividing one beat (one quarter note) into six, the eighth note is an eighth note. (One quarter note can be expressed in one beat). On the other hand, in the resolution of the sixteenth note system and the sixt note system, one beat (quarter note 1)
1 to 4 out of 12 pronunciation units
６6, 7９9, and 10 to 12 represent a sixteenth note (four sixteenth notes can be represented in one beat), and one to four, five, and twelve of the twelve pronunciation units. ~ 8,9 ~
In each of the twelve, each sound constituting a sixt is expressed (one sixt can be expressed in one beat).
Represents a quarter note in the pronunciation units 1 to 12 (1
One quarter note can be expressed in a beat). The tone generation timing and tone generation time of the musical tone are controlled for each of the six or twelve divided tone generation units.

When the musical tone data of each part is stored in the memory according to the above-described resolution, if the resolution is an eighth note system or a triplet system (QT0 to QT3 value "0"), One eighth note is stored using three storage locations (corresponding to 1 to 3 and 4 to 6 out of the above six pronunciation units), and two storage locations are used for triplets And memorize one note that makes up the triplet,
For quarter notes, six storage locations are used to store one quarter note. Further, in a sixteenth note system and a sixt note system (QT0 to QT3 have a value of "1"), the sixteenth note has three storage positions (1 to 3, 4 out of the twelve pronunciation units). , 6, 7-9, and 10-12), and stores one sixteenth note. For a sixt, one note that forms a sixt using two memories Remember minutes. As described later, the tone data stored for each resolution is read out in accordance with the resolution to control the tone generation, thereby generating a tone of a predetermined note. it can. In this way, by storing the musical sound data in a form corresponding to the resolution corresponding to the musical note, the storage capacity of the musical sound data memory can be reduced. That is, in the above example, the number of storage positions of musical tone data for one eighth note and one sixteenth note is only three, and the conventional common resolution (sixteenth note) Resolution), one eighth note can be converted to 6
The memory capacity is small as compared with the case where the data is stored using one storage location. By the way, in this embodiment, the above-described processing for storing musical tone data according to the resolution is employed for the base part, the chord 1 part, and the chord 2 part, and another method described below is employed for the rhythm part. Have been. That is, for the rhythm part, the musical sound data is stored by the event method described above. In the event method, the resolution is determined by a time length that can be expressed by the interval time, that is, the number of bits of the interval time data. In this embodiment, all the interval times of the tone data for a plurality of styles (rhythm types) relating to the rhythm part are expressed at a resolution corresponding to a common specific note, the tone data is stored, and the tone data read processing is performed. In this method, a musical sound (rhythm instrument sound) expressed at the original resolution is generated by performing time measurement corresponding to the interval time in accordance with the original resolution. Even in the case of such an event system, the number of bits of the interval time can be saved (because the length corresponding to the eighth note and the length corresponding to the sixteenth note can be expressed by the same value). As described above,
The resolution is set for each part in accordance with the tone data (note type) of each part, and the resolution (QT0 to QT) of each part is set.
According to T3), the tone data of the part is processed.

Next, the rhythm pattern area 9b will be described. As shown in FIG. 3A, the area RDT0 includes a data identification flag DTFLG, an instrument number NINST, an interval time INTTM, and an end code (0x0).
f (h)). First, the data identification flag DTFLG stores data indicating whether the data stored at the next address is the instrument number NINST or the interval time INTTM. Therefore, if the value of this data identification flag is data representing "the data stored at the next address is the instrument number" (see * 1, * 3, * 5 in the figure), The musical instrument number NINST is stored in the address. The musical instrument number NINST stores data representing the type of rhythm musical instrument, for example, a bass drum, a snare drum, a carbell, and the like. On the other hand, when the value of the data identification flag is data representing "the data stored at the next address is an interval time" (* 2, *
4, * 6), the interval time INTTM is stored at the next address. The interval time here refers to how many times the output process in step Sg5 of the rhythm output routine (described later) in FIG. 10 and the step Sg7 are performed before the next output process of the same step is performed. Means. Therefore, for monitoring the interval time, the register R
This is performed by decrementing the interval time set in the INTTM by "1".
The sounding timing can be controlled. The details of this process will be described later.

Next, the base pattern area 9c and the code pattern area 9d will be described. FIG. 3 (b),
(C) and (d) respectively show an area RDT1 representing a base part, an RDT2 representing a code 1 part, and a code 2
RDT3 representing a part is shown. In these areas, data represented by the difference key code ΔKC is stored in the order of pronunciation. Then, using the value of the difference key code ΔKC, a key code KC representing the pitch of the generated musical tone is determined. The key code KC is determined by “KC = ΔKC + BGN” using the value of the above-described pitch start point BGN. The value of the difference key code ΔKC is
“0” is a key-off code indicating a key-off, and “1” is a key-off code.
(H) to E (h) ”represent difference data (ie, pitch data) from the value of the pitch start point BGN representing the pitch start point;
“F (h)” represents an end code indicating the end of the performance data. Further, each region RDT1, RDT2, RDT
3, a plurality of data are set according to the style of the automatic performance in addition to the data shown.

B: Operation of the Embodiment The operation of this automatic performance apparatus will be described with reference to the flowcharts shown in FIGS. FIG. 4 shows a main routine, and the processes shown in FIGS.
This is a subroutine referred to in this main routine.
FIGS. 8 and 9 show a timer interrupt control routine. FIGS. 10, 11 and 7 show subroutines referred to in the timer interrupt control routine.

(1) Overview of Overall Operation of Automatic Performance Apparatus An overview of the overall operation of the automatic performance apparatus will be described with reference to a flowchart shown in FIG. First, when the power is turned on to the automatic performance apparatus shown in FIG. 1, the initialization shown in step Sa1 is performed. In this initialization, various registers are reset to zero, and the process proceeds to step Sa2.
In step Sa <b> 2, a process of determining whether or not a key press / release event of any key on the keyboard 1 has been detected by the key event detection unit 2. In this case, if a key processing event is detected, the determination result is “YES”,
The process proceeds to the sound generation / silence processing of the next step Sa3. In step Sa3, the tone generation process or the tone generation process is performed according to whether the key event is a key press event or a key release event.
A tone silencing process is performed, and the process proceeds to step Sa4. In the chord detection process of step Sa4, a process of setting chord detection data to be used during automatic performance is performed, and the detection result is stored in the registers RT and TP. In the process of step Sa4, a chord is detected from the key pressed state, and data indicating the root of the detected chord and the chord type are stored. Now, after the processing of step Sa4, or when it is determined “NO” in step Sa2, the processing shifts to processing of step Sa5.

Next, in step Sa5, an automatic performance switch (a switch in the switch group 3 shown in FIG. 1)
The on / off operation detection process is performed. If the on / off operation is detected and determined to be "YES", the process proceeds to a start / stop routine of step Sa6. In the start / stop routine of step Sa6, a start process or an end process of the automatic performance is performed by referring to a performance state flag RUN described later. By the way, after the processing of step Sa7 or in step 5, the ON / OFF operation of the automatic performance switch is not detected and "N
If it is determined to be "O", the flow proceeds to step Sa7.

Next, in step Sa7, a process of detecting the on / off operation of the performance style selection switch (the switch in the switch group 3 shown in FIG. 1) is performed. When the on / off operation is detected, the process proceeds to step Sa8. In step Sa8, a performance style setting routine, which will be described later, is performed by referring to the performance state flag RUN and the like. Thereafter, the flow returns to the key event detection processing in step Sa2 described above. On the other hand, if the on / off operation of the performance style selection switch is not detected in step Sa7, the process immediately returns to the key event detection process in step Sa2. As described above, after the power is turned on to the automatic performance device, after the initial setting is performed in step Sa1, the processing from step Sa2 to step Sa8 is repeatedly performed. Note that the performance state flag RUN
Is set based on the operation of the automatic performance switch. If the signal value of the performance state flag RUN is "0", it indicates "performance stop state", and if it is "1", it indicates "playing state". The value of the style change flag STSW is set based on the operation of the performance style selection switch. The style change flag STSW is set to "1" when the performance style selection switch operation is detected when the automatic performance device is in the "playing state", and is set to "0" otherwise. It is a register.

(2) Start / Stop Routine and Performance Style Setting Routine Next, the start / stop routine will be described in detail with reference to the flowchart of FIG. When the switch event detector 4 detects an operation of the automatic performance switch in the switch group 3 (see FIG. 1), the performance state flag RU
The inversion of the signal value of N is performed (step Sb1). Here, when the performance state flag RUN changes from the performance stop state value "0" to the playing state "1", the determination in step Sb2 becomes "NO", and the routine proceeds to the accompaniment initialization routine shown in step Sb4. .

The accompaniment initial setting routine performs an initial value setting process of various registers as shown in FIG. After turning on the power, various registers in this accompaniment initial setting process
If the performance style setting switch is not pressed, the default value is stored. If the switch is pressed, the value set in the performance style setting routine (see FIG. 6) is stored. Here, the performance style setting routine (see step Sa8 in FIG. 1) will be described with reference to FIG.

When a performance style selection switch (disposed on the switch group 3 shown in FIG. 1) is operated, first,
In step Sc1, a process of determining the value of the performance state flag RUN is performed. In this determination processing, if the performance state flag RUN is “1” indicating the “playing state”, the process proceeds to step Sc2, and the register NSTYL stores the selected style number corresponding to the operation state of the performance style selection switch. Set. Thereafter, the process proceeds to step Sc4. In step Sc4, the style change flag STSW set to “0” in the initial setting (see step Sa1 in FIG. 1) is set to “1”. Then, the process returns to the main routine. On the other hand, if the performance state flag RUN is "0" in the determination processing in step Sc1, that is, if the performance state flag RUN indicates the "performance stopped state", the register STY is set as shown in step Sc3.
Set the selected style number in L. Thereafter, the process exits the performance style setting routine.

Here, the functions of the registers NSTYL and STYL will be described. The selected style number stored in the style STYL is directly used for designating a performance style (see step Sd1 described later), but the register NS
The selected style number stored in TYL is the style ST
Used after being transferred to YL. This is because it may be unnatural to change the performance style immediately after the performance style selection switch is operated during the performance. Therefore, the selection style number is stored in the register NST.
The data is temporarily stored in YL, and is transferred to the register STYL and used at the transition of a bar or at the end of the current performance. The transfer process of the selected style number will be described later.

Returning to the accompaniment initialization routine of FIG. 7, first, in step Sd1, data in the style header area 9a corresponding to the value of the register STYL is read, and the read data is stored in the corresponding register. Store. That is, beat BT, tempo TMP tone VO
ICE1 to VOICE3, resolutions QT0 to QT3, and pitch start point BGN are respectively set in a register RTMP, registers RVOICE1 to VOICE3, and registers RQT0.
RQT3 and register RBGN are set. afterwards,
Proceed to step Sd2. In step Sd2, the register R is assigned to the channel corresponding to each performance part of the sound source.
The value (tone) set in VOICE1 to RVOICE3 is output. Thereby, the timbre of each channel is determined. Thereafter, the process proceeds to step Sd3. In step Sd3, registers DT0 to DT3, performance data storage addresses P0 to P3, counter Q, style change flag S
A process of writing all “0” to the TSW and the register RINTTM is performed. Then, the process goes out of the accompaniment initialization process,
The process exits the start / stop routine shown in FIG.

Here, the registers DT0 to DT3 will be described. After performing the rhythm output routine (step Se3 in FIG. 8) or the base / code output routine, the registers DT0 to DT3 store the “remaining time (equal to the number of remaining interrupts) ms” before performing the next routine.
"ec" for each performance part. Immediately after the automatic performance switch is pressed, "0" is set so that sound generation processing (rhythm output routine and base / chord output routine) is immediately performed.

On the other hand, in step Sb2 shown in FIG. 5, when the performance state flag RUN changes from the "playing state" to the "performance stopped state", that is, from "1" to "0", the processing in step Sb3 is performed. Proceed to. In step Sb3, an accompaniment channel silencing process of the sound source is performed to stop the sound emitting state. Thereafter, the process proceeds to step Sb5. In step Sb5, a determination process of the value of the style change flag STSW is performed. In this determination processing, when the style change flag STSW is "1", that is, when the performance style selection switch is operated during the "playing state", step Sb6 is performed.
After the selected style number set in the register NSTYL is assigned to the register STYL, the style change flag STSW is returned to "0". Now, step Sb
6 or in step Sb5, the style change flag STSW is set to “0”, that is, “N”.
If it is determined to be "O", the process exits the start / stop routine.

(3) Timer Interrupt Control Routine Next, a timer interrupt control routine started by an interrupt request of the timer 8 generated every 1 msec will be described with reference to FIGS. 8 to 11 (FIGS. 10 and 11 are subroutines). explain. Note that the flowchart shown in FIG. 8 shows the performance output timing control processing of each performance part, and the flowchart shown in FIG. 9 shows the processing for bar monitoring control processing. Now,
The rhythm output routine (step Se3) and the bass / chord output routine (step Se8) referred to in the performance output timing control process shown in FIG.
"Actual pronunciation processing" is performed. So, first,
The rhythm output routine (step Se3) and the base / chord output routine (step Se8) will be described.

Rhythm Output Routine and Base / Chord Output Routine a. Rhythm Output Routine (See FIG. 10) First, this rhythm output routine is referred to during the first interrupt processing immediately after the automatic performance start switch is pressed. Will be described. The determination result of the value of the register RINTTM in step Sg1 is “YES”.
become. This is because the register RINTTM has been initialized to "0" in advance by the accompaniment initialization routine shown in FIG. Next, the process proceeds to step Sg2. The result of the determination in step Sg1 is “NO”.
The case will be described later. Next, in step Sg2, it is determined whether or not the value of the region RDT0 (STYL, P0) referred to according to the value of the style STYL is “end code (F (h))”. The result of this determination is "NO". This is because, as described above, immediately after the automatic performance start switch is pressed, the automatic performance starts from this point. Then, step Sg3
Proceed to.

In step Sg3, the region RDT0
A determination process is performed to determine whether or not the data identification flag DTFLG stored in (STYL, P0) is “instrument code”. The result of this determination is "YES". This is because the data arrangement is as shown in FIG. Then, the process proceeds to the next step Sg4. In step Sg4, the region RDT0 (STYL, P0 + 1),
The musical instrument number NINST stored at the address next to the area RDT0 (STYL, P0) is stored in the register RNINST.
Set to. Thereafter, the process proceeds to step Sg5.

In step Sg5, the key-on signal K
ON and the value set in the register RNINST are output to the rhythm channel of the sound source. As a result, the tone generator 10 (see FIG. 1) performs musical tone generation processing of the rhythm instrument sound corresponding to the designated instrument number, and the tone is generated by the sound system 11 (see FIG. 1). Next, the process proceeds to step Sg6. Step Sg6
In, the processing for setting the address at which the next performance data is stored in the performance data storage address P0 is performed. In this case, as shown by “P0 = P0 + 2”,
A value obtained by adding "2" to the performance data storage address P0 is set. This is, as shown in FIG. 3A, "data identification flag DTFLG and musical instrument number NINST" and "data identification flag DTFLG and interval time INT.
This is because the "TM" and the "TM" are stored as a pair using an "area for two addresses". Thereafter, the process returns to step Sg2.

In step Sg2, the style STYL is read from the newly set performance data storage address P0.
A determination process is performed to determine whether or not the region RDT0 (STYL, P0) referred to in response to the end code is the end code (F (h)). In this case, as described above, since it is immediately after the sound data is output in step Sg5, and the value of the register RINTTM for setting the interval time needs to be reset, the determination result is "NO", and the determination result is "NO".
Proceed to g3. In step Sg3, the region RDT0
The data stored in (STYL, P0 + 1) is
Since it is "not an instrument code", it is determined to be "NO" and the process proceeds to step Sg8. In step Sg8,
The interval time INTTM stored in the area RDT0 (RP + 1) is set in the register RINTTM. Here, (RP + 1) is
This indicates that the address is next to the area RDT0 (STYL, P0) (the same as P0 + 1). Next, the process proceeds to step Sg9. In step Sg9, a process of setting the address where the next performance data is stored in the performance data storage address P0 is performed (see step Sg6). In this case, as shown in the figure, a value obtained by adding "2" to the performance data storage address P0 is set.
Thereafter, the process exits the rhythm output routine.

Next, it is assumed that a second interrupt request occurs after the automatic performance start switch is pressed, and the interrupt control routine shown in FIG. 8 has advanced to the rhythm output routine of step Se3. In this case, the determination process in step Sg1 of the rhythm output routine shown in FIG.
Since it is immediately after the new value is set in the register RINTTM, "NO" is determined, and the process proceeds to step Sg7.
In step Sg7, a process of decrementing the value of "1" from the register RINTTM is performed, and the process exits the rhythm output routine.

Thereafter, every time an interrupt request of the timer 8 is generated, the register RINTTM is decremented. Then, when the register RINTTM becomes "0", steps Sg2, Sg3, and Sg4 are performed in the same manner as when the first interrupt request occurs.
In step Sg5, a data (NINST, KON) tone output event for generating a rhythm instrument sound is output to the tone generator 10 (FIG. 1). As a result, the tone generator 10, the sound system 11, and the like perform the sounding processing of the next rhythm instrument sound.

As described above, the rhythm part automatic performance processing is performed by the above-described performance data output processing (step Sg5).
And sounding timing count processing (step Sg7)
Is appropriately and selectively performed based on the value of the register RINTTM.

Next, a description will be given of a process in which the automatic performance process proceeds to produce the last sound of the music. In this case, after “YES” is determined in step Sg1, step Sg2, step Sg3, step Sg4
In step Sg5, the rhythm instrument sound pronunciation data (NINST, KO
N) is output to the tone generator 10 (FIG. 1). Thereby, the tone generator 10
The sound system 11 and the like perform sound processing of the last single sound. Next, the process proceeds to step Sg6, in which the performance data storage address P0 is set to the address where the next performance data is stored. Then, step Sg2
Return to

In step Sg2, the style STYL is read from the newly set performance data storage address P0.
RDT0 (STYL, P0) referenced according to
Is an end code (0x0f (h)). In this case, as described above, immediately after the last note is generated in step Sg5, the end code (0) is stored in the area RDT0 (STYL, P0).
x0f (h)), the judgment result is “Y
ES ", and immediately exits the rhythm output routine.

B. Base / Code Output Routine (See FIG. 11) Next, the code / base output routine will be described.
First, a description will be given of a case where this chord / bass output routine is referred to during the interrupt processing when the last note of the music is generated from the first interrupt processing immediately after the automatic performance start switch is pressed. . In the first state, the value of the register i is set to “1”.
In step Sh1, the area PDTi (STYL, STYL,
Pi) (since i = 1, hereinafter, until the register i changes, a process of determining whether or not the region RDT1 (represented as STYL, P1)) is an end code (0x0f (h)). Immediately after the automatic performance start switch is pressed, and since the data arrangement is as shown in FIGS. 3B to 3D, "NO" is determined and the process proceeds to step Sh2.
In step Sh2, the region RDT1 (STYL,
The data stored in P1) is the key-off code (0x
00 (h)). Immediately after the automatic performance start switch is pressed, any case may occur.

Therefore, in step Sh2, "Y
If it is determined to be "ES", the process proceeds to Step Sh4.
In step Sh4, the key-off signal KOFF is output to the channel corresponding to the base part (i = 1) of the sound source. Thereafter, the process proceeds to Step Sh7. On the other hand, if “NO” is determined in the step Sh2, the process proceeds to a step Sh3. In step Sh3, the value of the pitch start point BGN and the area PDT1 (STY
(L, P1 + 1) is set. Thereafter, the process proceeds to Step Sh5. In step Sh5, the key code KC set as described above is detected. In this case, in the chord detection processing (see step Sa4 in FIG. 4), the set register RT
And a pitch conversion process based on the value of the register TP. Thereafter, the process proceeds to Step Sh6. In step Sh6, the key-on signal KON and the key code KC are output to the channel corresponding to the base part (register i = 1) of the sound source, and the flow advances to step Sh7.

In step Sh7, the performance data storage address P1 is incremented by "1" by "P1 = P1 + 1", and the next referenced area RDT1 (STY
(L, P1). Then, after the processing of step 7, the process exits the base / code output routine. The processing of Step Sh1 to Step Sh7 is performed in the same manner as described above even when the value of the register i corresponds to “2” and “3”, that is, the code 1 part and the code 2 part. Then, the above-described processing is performed for each performance part every time the interrupt control routine is started until the last sound of the music is pronounced.

Next, a case will be described in which reference is made during interrupt processing immediately after the last sound of a musical piece has been generated. In this case, the area PDT1 (STYL, P1)
Since the end code is stored, the result of the determination processing in step Sh1 is “YES”, and the process immediately exits the base / code routine.

Each performance part output timing control process In the performance part output timing control process shown in FIG. 8, "sound generation timing control according to resolution" is performed together with the calculation process of "division period for quarter note". ing. First, in step Se1, it is determined whether or not the performance state flag RUN is "1", that is, whether or not the "performance state". The result of this determination processing is “N
If it is "O", that is, "performance stopped state", the process immediately exits the timer interrupt control routine and returns to the main routine. On the other hand, if the determination processing in step Se1 is “YE
S ", that is, the" playing state ", the process proceeds to the next step Se2.

In the processing from step Se2 to step Se5, the sounding timing of the rhythm part is controlled in accordance with the value of the resolution QT0 of the rhythm part. First, the details of the process of determining the value of the register DT0 in step Se2 will be described. When an on event of the automatic performance switch (disposed in the switch group 3) is detected by the switch event detection unit 4, the performance state flag R
UN is set to "1", that is, "playing state" (step Sb1 in FIG. 5), and the value of the register DT0 is initialized to "0" (step Sd3 in FIG. 7).
In this state, the interrupt request of the timer 8 is
Is detected, the process proceeds to step Se2 via the determination process of step Se1. In this step Se2,
A determination process is performed to determine whether or not the relationship of “DT0 <1/2 (msec)” is satisfied. In this case, as described above,
Since the automatic performance switch-on event has just been detected, the value of the register DT0 is “0”. Therefore, the result of this determination is "YES", and the routine proceeds to the rhythm output routine of step Se3. Then, after performing the rhythm output routine shown in step Se3 in the first interrupt processing, the process proceeds to step Se5.

In step Se5, in the timer interrupt control repeatedly performed every 1 msec, what is the "remaining time (remaining interrupt count) msec" until the rhythm output routine (step Se3) in the second interrupt processing is performed. The remaining is calculated by the following formula.

[0046]

[Expression 1] “DT0 = (60 × 1000) / RTMP × 1 / ((R
QT0 + 1) × 6) -1 + DT0 (msec) "

In the equation (1), (60 × 1000 / R
(TMP) (msec) represents the sounding time of a quarter note according to the value of the tempo TMP (read out to the register RTMP). Next, 1 / ((RQT0 + 1) ×
6) (Dimensionless) is the value of the resolution QT0 (register RQT
(Read to 0) indicates the division ratio of the sounding time of the quarter note.

Therefore, (60 × 1000) / RTM
P × 1 / ((RQT0 + 1) × 6) (msec) is a “division period for quarter note” when a quarter note is divided by “a predetermined number of divisions” represented by ((RQT0 + 1) × 6). ". This "division period for quarter notes"
Is expressed in units of milliseconds, and can be counted in synchronization with the cycle (1 msec) of the interrupt request of the timer 8. Also, in the first step Se5, “1”, that is, “1 msec” is decremented because the first rhythm output routine is performed, so that the “division cycle for quarter note” is changed to “1 ms”.
ec ”indicates consumption. In other words, the "remaining time (remaining interrupt count) msec" until the second rhythm output routine is performed is "1 msec", indicating that the time has decreased, and an accurate value of how many msec to measure is set. doing. The reason why “DT0” is added is to reduce errors in tone generation timing control for each performance part. The details will be described later.

On the other hand, the judgment processing in step Se2 is "NO", that is, "DT0 <1/2 (ms
ec) ”does not satisfy the relationship of step Se.
Proceed to 4. In step Se4, step Se described above is performed.
The value of the register DT0 in which the “divided cycle for quarter note” set in step 5 is set, and thus the “remaining time (remaining interrupt count) msec” is set to “1”, that is, “1 msec” for each interrupt request of the timer 8. Decrement by one. Thereby, in step Se2, “YE
The time until it is determined as “S”, that is, “the division cycle setting timing for the quarter note” is measured.

The "divided cycle setting timing for a quarter note" is determined by the determination processing in step Se2. In this determination processing, the register DT
"Remaining time (remaining interrupt count) m" set to 0
"sec" is determined by the determination threshold value. The reason why the judgment threshold value is provided is that the “division period for quarter notes” calculated in step Se5 is determined for each performance part.
Because it is slightly different. The fact that "division periods for quarter notes" are different means, strictly speaking, "the sounding timing is different". In this case, since a certain degree of error cannot be denied, if the deviation of the sounding timing of each performance part falls within "1/2 (msec)", the sound is generated. Therefore, this threshold does not have to be limited to “１ ／ msec”. That is, if accuracy is required, the threshold value may be reduced, and if roughness is sufficient, the threshold value may be increased.

The process from step Se6 to step Se12 is a process for each performance part, that is, a process for controlling the output timing of each part of the bass, chord 1 and chord 2, and the process in steps Se2 to Se5 described above. According to the processing. Each performance part corresponds to the value of the register i. That is, the possible values “1”, “2” and “3” of the register i are:
Each corresponds to a bass, chord 1 and chord 2 parts. First, in step Se6, "1" is set in the register i. As a result, in the subsequent processing from step Se7 to step Se11,
The output timing control of the base part is performed.

In step Se7, it is determined whether or not the value of the register DTi (i = 1) is smaller than "1/2 (msec)". This determination process is "YE
S ", that is, if the relationship of" DTi <1/2 (msec) "is satisfied, the base / code output routine of step Se8 is performed. In the base / chord output routine of step Se8, a performance part corresponding to the value of the register i is output. That is, in this case, the performance part of the bass part is output (i = 1). When this processing ends, the register D shown in the step Se10 is read.
The process proceeds to the calculation of Ti. In this process, the value of the register DTi (i = 1) is calculated according to the value of the register i in the same manner as in step Se5. Next, the process proceeds to step Se11.

At step Se11, the register i
Is incremented, and processing for specifying the next part is performed. Then, the flow shifts to step Se12, where it is determined whether the value of the register i is "4". If this determination is "NO", the flow returns to step Se7, and the above-described processing is repeated again. This repetition is performed until the value of the register i becomes "4". As a result, the same processing as described above is performed for all parts. Note that the processing performed in steps Se7 to Se10 is the same as the processing in step Se2 described above.
To Step Se5. Thereafter, the processing for all parts is completed, and the value of the register i is set to “4”.
Then, the determination in step Se12 becomes “YES”, and the flow proceeds to the measure monitoring control process shown in FIG.

Now, regarding the above-described processing, specifically, the case where musical tone data of a rhythm part composed of a note longer than a triplet and a note longer than a sixt is generated. Will be described. If tone data composed of notes longer than the triplet is generated, the value of the register DT0 is calculated in step Se5 using the resolution QT0 set to "0". As described above, when the resolution QT0 is “0”, as described above, the musical tone data for one quarter note is divided and stored in the six storage areas of the rhythm pattern area 9b in the ROM 9. Therefore, by executing the rhythm output routine of step Se3 six times, one quarter note is generated. Therefore, the eighth note can be sounded by performing the rhythm output routine of step Se3 three times, and the triplet can be sounded by performing the rhythm output routine of step Se3 twice.

If tone data composed of notes longer than the length of the sixt is to be generated, the value of the register DT0 is determined in step Se5 using the resolution QT0 set to "1". calculate. As described above, when the value of the resolution QT0 is “1”, the musical tone data for one quarter note is divided into twelve pieces and stored, so the rhythm output routine of step Se3 is performed 12 times. By doing so, one quarter note is pronounced. Therefore,
The eighth note, the sixteenth note, and the sixth triplet can generate a corresponding note by executing the rhythm output routine of step Se3 six times, three times, and twice, respectively.

The same applies to the case where the bass / chord part is sounded. That is, according to the type of the note and the values of the resolutions QT1 to QT3, the base / chord part shown in step Se8 is performed several times to generate one note. In step Se5 and step Se10, "DT0" (DTi) is added in these steps as described above, because the "division period for quarter note" in the rhythm part set in each step and each performance part This is because the value is different from the “division period for the quarter note”. That is, as a result, the sounding timing of each performance part is slightly shifted. The deviation of this sounding timing is "1 / 2mse
If the value is less than "c", the value is used to determine the next sounding timing (that is, added) to reduce the error.

Measure monitoring control processing Next, the measure monitoring control processing will be described with reference to FIG. First, in step Sf1, the timer interrupt counter Q is incremented by "1". For example,
If it is the first interrupt request immediately after the automatic performance start switch is pressed, the timer interrupt counter Q is initialized to "0" by the accompaniment initialization routine shown in FIG.
Is set to “1”. Next, the process proceeds to step Sf2.
In step Sf2, the register RBT (beat B
(T is written) is "0", that is, whether or not it is a quarter beat. In this case, if "YES", a value of "4" is set in the register CNT (step Sf3), and if "NO", a value of "3" is set in the register CNT (step Sf4). Step S
After the processing of f3 or step Sf4, step Sf
Go to 5.

In step Sf5, the timer interrupt counter Q counts the number of timer interrupts ((60
X1000 / TMP) × CNT × 2) (dimensionless) is determined. The number of timer interrupts ((60 × 1000 / TMP) × CNT × 2) is
Since the timer 8 issues an interrupt request every 1 msec,
It also represents the performance time for two measures. Well, first,
The case where this determination processing is “NO”, that is, the case where they are not equal will be described. If the answer is "NO",
This is the case where the performance location is not a two-measure break. In that case, the process first proceeds to step Sf6. Next step Sf6
In, the timer interrupt counter Q is set to the number of timer interrupts for one bar (60 × 1000 / TMP × CNT)
(No dimension) is determined. If the determination is "NO", that is, if the performance location is in the middle of a bar, the process immediately exits the timer interrupt control routine. On the other hand, if this determination is "YES", that is, if the performance location is at a one-measure break, the process proceeds to the next step Sf7.

In step Sf7, a process is performed to determine whether or not the style change flag STSW is "1", that is, whether or not the performance style change switch has been operated in the "playing state". If the determination in step Sf7 is "NO", that is, if the performance style change switch has not been operated, the process immediately exits the timer interrupt control routine. On the other hand, if the determination in step Sf7 is “YES”, the process proceeds to step Sf7.
Proceeding to step 8, the processing for silencing the accompaniment channel of the sound source is performed. Next, the process proceeds to step Sf9. Step Sf9
In, the processing of setting the value of the register NSTYL holding the parameter specified based on the above-described performance style change switch operation to the style STYL is performed. Then, the process proceeds to the next Step Sf10. Step S
In f10, an accompaniment initial setting process (see FIG. 7) is performed, and the style STY set in step Sf9 is performed.
Processing for resetting various parameters according to L is performed.
Thereafter, the process exits the timer interrupt control routine.

Next, the case where the result of the judgment processing in step Sf5 described above is equal, that is, the case where the performance location is a break in units of two measures will be described. In this case, the process proceeds to step Sf11, and a determination process is performed to determine whether or not a performance style change switch operation has been performed. If the determinations are equal, the process proceeds to step Sf8 described above. Then, after performing the processing of step Sf8, step Sf9, and step Sf10 in order, the process exits the timer interrupt control routine. On the other hand, step Sf1
If the determination processing in step 1 is not equal, that is, if the performance style change switch has not been pressed, the process proceeds to step Sf12. Step Sf
In step 12, the values of the registers DT0 to DT3, the timer interrupt counter Q, the performance data storage addresses P0 to P3, and the register RINTTM are all initialized to "0". Thereafter, the process exits the timer interrupt control routine.

As described above, the measure monitoring control processing is performed in two steps.
The repetition is performed in measures, and the change of the performance style is performed in measures.

(4) Overall Operation In the above embodiment, the tone generation timing is controlled in a time-sharing manner by utilizing the fact that an interrupt request by the timer 8 is generated every 1 msec. That is, the register RTM
Using parameters such as P, the register RINTTM, the resolutions QT0 to QT3, and the timer interrupt counter Q, it is possible to perform processing corresponding to each part, each measure, and each note type. In this case, the actual tone generation timing control processing is all performed in this interrupt processing.

In this case, in particular, the output timing of the musical tone is monitored for each interrupt control request by the processing shown in steps Se2 to Se5 in FIG. 8 (the same applies to steps Se7 to Se10). In addition, since the setting of "division period for quarter note" is performed for each tone output timing, the tone generation timing of the tone can be accurately monitored according to the value stored in the resolution QT0 (QTi). . Further, since the value of the register RTMP is reset at the same time, the speed of the musical sound performance can be changed in response to the tempo changing operation in real time. Also, in step Se2, "1/2 mse
By using "c" as the determination value, it is possible to prevent accumulation of errors in sounding timing due to "division periods for quarter notes" having different values for each rhythm part and performance part.

[0064]

As described above, according to the present invention,
To generate a musical tone corresponding to a plurality of types i.e. the automatic performance of the performance data comprising a plurality of parts, instead of setting the resolution to match the part requiring the most fine resolution in common, "each part Set the “optimal resolution” individually. Therefore, there is no need to use a useless memory as in the past, and the circuit scale can be reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an overall configuration of an automatic performance device according to an embodiment of the present invention.

FIG. 2 is a memory layout diagram showing a mapping state of a style header 9a in the embodiment.

FIG. 3 is a memory layout diagram showing a mapping state of each part performance data (rhythm pattern area 9b, base pattern area 9c, chord pattern area 9d) in the embodiment.

FIG. 4 is a flowchart showing processing of a main routine.

FIG. 5 is a flowchart showing a process of a start / stop routine.

FIG. 6 is a flowchart showing the processing of a performance style setting routine.

FIG. 7 is a flowchart showing processing of an accompaniment initialization routine.

FIG. 8 is a flowchart showing a part output timing control process performed in a timer interrupt control routine.

FIG. 9 is a flowchart showing a bar monitoring control process performed in a timer interrupt control routine.

FIG. 10 is a flowchart showing a rhythm output routine.

FIG. 11 is a flowchart showing a base / code output routine.

[Explanation of symbols]

7 CPU 9a Style header area 9b
Rhythm pattern area, 9c ... Base pattern area, 9
d: Code pattern area.

Claims (2)

(57) [Claims] 1. A sequentially reads the previously stored performance data, the automatic performance apparatus for performing an automatic performance in accordance with the read data, storage means for storing the performance data comprising a plurality of parts to be read simultaneously, before Symbol performance data the corresponding resolution data to determine the resolution of the sounding timing and sounding time of the tone, and resolution setting means for setting for each of the <br/> each part, based on the resolution data of the each part, each corresponding Pas
An automatic performance device comprising: reading control means for reading out performance data of a music piece in parallel at a cycle corresponding to the resolution . 2. A sequentially reads the previously stored performance data, the automatic performance apparatus for performing an automatic performance in accordance with the read data, storage means for storing the performance data comprising a plurality of parts to be read simultaneously, before Symbol performance data The corresponding musical tone
Resolution data that determines the resolution of sounding timing and sounding time
A resolution setting unit for setting the over data in each part, the period a total according to the resolution data set in each part
A plurality of counting means for performing a time, based on time measurement timing at which the counted by the counting means
Zui, the automatic playing apparatus characterized by comprising a read control means for reading the performance data of each corresponding part.