JPH0738112B2 - Automatic musical instrument accompaniment device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic accompaniment apparatus for an electronic musical instrument, and more particularly to improvement of an automatic accompaniment apparatus of a type that sequentially reads pattern data according to sequence data.

[Summary of the Invention] The present invention enables automatic accompaniment of various types with a small storage capacity by storing sequence data and a plurality of pattern data for each chord type group in the automatic accompaniment apparatus of the type described above. By adjusting the pitch of the accompaniment sound signal according to the pitch correction data, the generation of dissonance is prevented.

[Prior Art] Conventionally, as an automatic accompaniment device for electronic musical instruments, sequence data and a plurality of pattern data are stored for each chord type, and the chord type and the chord root note are stored based on the key depression state on the accompaniment keyboard. Is detected, the sequence data and a plurality of pattern data corresponding to the chord type related to the detection are selected, and the plurality of pattern data related to the selection are sequentially read out according to the sequence data related to the selection, whereby the pitch in the pattern data is detected. It is known to generate an accompaniment tone signal having a pitch determined by the data and the chord root to be detected (see, for example, JP-A-61-158400).

[Problems to be Solved by the Invention] According to the above-mentioned conventional technique, for example, when handling a long accompaniment sound generation pattern of 8 bars, sequence data for designating a pattern for each bar of 8 bars and an object to be specified, for example, Since it is sufficient to store the pattern data for 3 bars, 8
Compared to storing pattern data for a measure, a small storage capacity enables automatic accompaniment with a wide variety of variations. But,
It is desired to further reduce the storage capacity.

In order to meet such a demand, it has been proposed to divide a large number of chord types into a plurality of groups and store sequence data and a plurality of pattern data for each chord type group. In this case, it has been found that the accompaniment sound of a specific note name is pronounced as a dissonance for a specific chord type.

It is an object of the present invention to further reduce the storage capacity and prevent the generation of discord.

[Means for Solving Problems] An automatic accompaniment device for an electronic musical instrument according to the present invention includes (a) an accompaniment and a keyboard, and (b) a chord type and a chord root note based on a key depression state on the keyboard. And (c) first storage means for storing accompaniment data for each chord type group out of a plurality of predetermined chord type groups, the accompaniment corresponding to each chord type group. The data includes a plurality of M pattern data each of which represents an accompaniment sound generation pattern of a performance section of a fixed length by an array of pitch data, and the first to Nth N-th data which specify these pattern data in an appropriate combination.
(N is an integer larger than M), and (d) determination means for determining which of the plurality of chord type groups the chord type detected by the detection means belongs to, e) selecting means for selecting in the first storage means accompaniment data corresponding to the chord type group determined to belong to this determining means; and (g) chord types belonging to the plurality of chord type groups. Second storage means for storing pitch correction data for preventing dissonance in correspondence with a predetermined chord type and a predetermined note name; and (h) sequentially according to sequence data in the accompaniment data selected by the selection means. By accompaniment sound generating means for generating an accompaniment sound signal having a pitch determined by the pitch data in the pattern data and the chord root detected by the detecting means by reading the pattern data in the accompaniment data. If the predetermined chord type is detected by the detection means and the pitch data of the predetermined pitch name is read from the first storage means, the pitch correction data of the second storage means is read. Accordingly, the pitch of the accompaniment sound signal is corrected so as not to cause dissonance.

[Operation] According to the structure of the present invention, when the detecting means detects the predetermined chord type and the pitch data of the predetermined pitch name is read from the first storing means, the accompaniment sound generating means outputs the second pitch. The pitch of the accompaniment person is corrected according to the pitch correction data of the storage means so that a dissonance does not occur.

[Embodiment] FIG. 1 shows a circuit configuration of an electronic musical instrument provided with an automatic accompaniment apparatus according to an embodiment of the present invention. In this electronic musical instrument, musical tones such as melody tones, chords and rhythm tones are produced. Is controlled by a microcomputer.

Circuit configuration (Fig. 1) Bus 10 has a melody keyboard circuit 12 and an accompaniment keyboard circuit 1
4, control switch group 16, clock generator 18, central processing unit (CPU) 20, program memory 22, accompaniment data memory 2
4, a register group 26, a tone generator (TG) 28, etc. are connected.

The melody keyboard circuit 12 includes a melody keyboard such as an upper keyboard, and key operation information is detected for each key.

The accompaniment keyboard circuit 14 includes an accompaniment keyboard such as a lower keyboard, and key operation information is detected for each key.

The control switch group 16 includes various control switches for tone control and performance control. Control switches related to the implementation of the present invention include a rhythm type selection switch and a rhythm start / stop switch. .

The clock generator 18 generates a clock signal,
This clock signal is used as an interrupt command signal to start the clock interrupt routine of FIG.

The CPU 20 executes various processes for generating musical tones according to the program stored in the program memory 22 composed of a ROM (Read Only Memory). For these processes, refer to FIG. 11 to FIG. See below.

The accompaniment data memory 24 includes a pattern ROM and a table ROM, and the pattern ROM stores accompaniment pattern information as described below with reference to FIGS. 7 to 10.
The table ROM stores various table information as described later with reference to FIGS. 3 to 6.

The register group 26 includes a large number of registers used in various processes by the CPU 20, and registers related to the implementation of the present invention will be described later.

The TG28 generates accompaniment sound signals such as chords, and has four sound generation channels as an example, and can sound up to four sounds at the same time. As the tone generator, in addition to the TG28, those for melody sounds, those for rhythm sounds, etc. are provided, but they are not shown.

The sound system 30 is for converting the musical tone signal from the tone generator into sound in the TG 28 or the like.

In the electronic musical instrument described above, the key pressing data corresponding to the key pressed on the keyboard and the pitch data forming the accompaniment pattern consist of key codes representing pitches, and the key code value is the pitch of each pitch. Each of them is predetermined as shown in FIG.

Contents stored in table ROM (FIGS. 3 to 6) The table ROM in the accompaniment data memory 24 includes a chord detection table CHDTBL (FIG. 3) and a group number table GRPTBL.
(FIG. 4), a pitch correction table PMTBL (FIG. 5), a chord progression determination table SEQREF (FIG. 6) and the like are included.

The chord detection table CHDTBL is used to detect chords based on the key depression state on the accompaniment keyboard.It stores chord data for each chord name and the chord data corresponding to the lowest note in the key depression. ing. Each chord data is 8-bit data, and the upper 4 bits represent a chord type and the lower 4 bits represent a root note.

As an example, the chord name with the root note as C has the stored contents as shown in FIG. 3, and, for example, the chord data (root note data and root note data representing C _M based on the key depression of CEG). The value of the chord type data is 0). Note that in FIG. 3, the key names shown in parentheses may omit the key depression.

FIG. 3 shows a storage unit for the root note C, but storage units similar to this are also provided for root notes other than C, respectively.

In chord detection, the lowest note that is _pressed is taken into consideration because different chords can be detected for chords with the same key depression state, such as C _m7 and E ^b ₆ , C ₆ and A _m7, etc. Is. Thus, for example if you depressed the ^{C-E b -G-B b} , the lowest tone is C, G, if either B ^b, chord data is obtained representing a C _m7, lowest note in E ^b If so, chord data representing E ^b ₆ is obtained.

The group number table GRPTBL is for discriminating the chord type group to which each chord type belongs. As an example, the group number GNO corresponding to the chord type group is stored as shown in FIG. In this example, M, M ₇ , 6
Is a major chord type group and is assigned a group number 0, m, m ₇ and m ₇ ^-5 are minor chord type groups and is assigned a group number 1 and 7, 7 _SUS 4 is a 7th chord type group The group number 2 is assigned to this.

In Fig. 4, the "TYPE" column is the chord type register TY.
The value of PE is shown. The chord type data stored in this register TYPE is the chord type M, M ₇ , 6, m, m ₇ , m ₇ ,
0, 1, 2, 3, 4, corresponding to 7, 7 _SUS 4 respectively
It takes a value of 5, 6, or 7. Chord type register TY
By referring to the group number table GRPTBL using PE, the group number (chord type group) corresponding to the value (chord type) of the register TYPE can be obtained.

The pitch correction table PMTBL is provided to correct the pitch data (key code) read from the pattern ROM so that no dissonant accompaniment sound is generated when the pitch is changed. As shown in FIG. 5, 0, -1,
The correction data having a value such as -2 is stored.

In FIG. 5, the “TYPE” column shows the value of the chord type register TYPE, as in the case of FIG. Also, "RKCNT"
Values of 4, 7, 9 and the like of the semitone register RKCNT are shown in the section of, and note names of E, G, A and the like corresponding to each value are shown in parentheses. Register ROM RKCNT contains pattern ROM
The semitone number data corresponding to the note name of the key code read from is stored. The number of semitones represents the number of semitones from the root note C, and takes any value of 0, 1 ... 11 corresponding to C, C ^# ... B, respectively.

The correction data to be read from the pitch correction table PMTBL is determined by the value of the register TYPE and the value of the register RKCNT.

The chord progression determination table SEQREF is provided to determine the transition mode to a specific chord (seventh in this embodiment). As an example, as shown in FIG. 6, 12 storage areas SEQREF _i (i = 0 to 11), and 8 for each storage area
It stores bit criterion data. Of the 8 bits of the judgment reference data, the lower 4 bits are pitch difference data representing the root pitch difference between the specific chord and the immediately preceding chord, and the upper 4 bits are the chord type immediately before the specific chord. This is the old chord type data to be represented. FIG. 6 illustrates, as an example, an old chord (a chord immediately before G ₇ ) corresponding to the determination reference data in each storage area when the current chord is G ₇ .

In FIG. 6, the column "SEVPT" shows the value of the pattern number register SEVPT for seven. This register S
The Seventh pattern number data stored in the EVPT has a value of 0, 1 or 2 corresponding to the Seventh normal pattern and the Seventh zeroth and first variation patterns, respectively. Here, the 0th variation pattern is a final accompaniment pattern,
The first variation pattern is a transitional accompaniment pattern.

A different value is set in the register SEVPT depending on what the chord immediately before the particular chord (Sevens) is. That is, when the immediately preceding chord corresponds to any of the judgment reference data in the storage areas SEQREF _{0 to} SEQREF ₇ , the register SEVPT
Is set to 1 and any of the judgment reference data of the storage areas SEQREF _{8 to} SEQREF ₁₁ is met, the register SEVPT
2 is set to _{0 and the} register SEVPT is set to 0 when none of the judgment reference data of SEQREF _{0 to} SEQREF ₁₁ is met.

Storage Contents of Pattern ROM (FIGS. 7 to 10) The pattern ROM in the accompaniment pattern memory 24 stores data as shown in FIG. 7 for each rhythm type such as march and waltz. In other words, FIG. 7 shows the data format corresponding to a specific rhythm type in the stored contents of the pattern ROM.

In FIG. 7, SEQ is a sequence storage unit, PTN is a pattern storage unit, ADRS is a first address storage unit, and ADRS2 is a second storage unit.
Address storage unit, CHASS is the first pronunciation number storage unit, CHASS
2 is the second pronunciation number storage unit, UPLMT is the upper limit pitch storage unit, QUAN
T is a tempo clock number storage unit.

The sequence storage unit SEQ includes the 0th, 1st and 2nd sequence storage units SEQ (0), SEQ (1) and SEQ respectively corresponding to major, minor and seventh chord types.
(2) is included. All of the 0th to 2nd sequence storages store the pattern progression for eight measures in the pattern number PNO.
The 16-bit sequence data represented by the array is stored, and each pattern number is represented by 2 bits.

As an example, the 0th sequence storage unit SEQ (0) stores “012
Sequence data representing a pattern progression such as “30123” is stored, and “0” is stored in the first sequence storage unit SEQ (1).
Sequence data indicating a pattern progression such as “1010101” is stored, and “0121201” is stored in the second sequence storage unit.
Sequence data such as "2" indicating the pattern progression is stored.

The pattern storage unit PTN includes 0th, 1st and 2nd normal pattern storage units PTNM, PTNm and PTNS respectively corresponding to major, minor and seventh chord type groups, and a variation pattern storage unit PTNV. I'm out.

The 0th normal pattern storage unit PTNM stores pattern data representing 0th to 3rd normal patterns for one bar each designated by the sequence data of the 0th sequence storage unit SEQ (0).

The first normal pattern storage unit PTNm stores pattern data representing 0th and 1st normal patterns for one bar each designated by the sequence data of the first sequence storage unit SEQ (1). The storage unit PTNm can further store an accompaniment pattern for two measures if necessary.

The second normal pattern storage unit PTNS stores pattern data representing the 0th to second normal patterns for one bar each designated by the sequence data of the second sequence storage unit SEQ (2). The accompaniment pattern for one bar can be further stored in the storage unit PTNS if necessary.

The variation pattern storage unit PTNV stores pattern data representing the 0th and 1st variation patterns for one bar, respectively. The 0th and 1st variation patterns are for the 7th, and the pattern of the storage unit PTNS is used as the 7th normal pattern. When the 7th chord is specified on the accompaniment keyboard, which of the normal pattern, the 0th and the 1st variation pattern is to be used is determined by the value of the register SEVPT (the pattern number for the 7th) and the SEVPT.
If is 0, the normal pattern is selected, if 1, the 0th variation pattern is selected, and if 2, the first variation pattern is selected. Therefore, with regard to the Seventh, a variety of automatic accompaniments can be performed by using three types of patterns.

When the sequence data and the pattern data are separately stored as described above and the pattern is selected and read out for each measure by the sequence data, when accompaniment patterns for eight measures are directly stored and read out in order of measures. In comparison, the amount of data to be stored is small. This is because, for example, even if the same pattern appears multiple times in 8 bars, only one pattern needs to be stored. Further, since the sequence data and the pattern data are stored for each chord type group such as major type, minor type, and seventh type, the amount of data to be stored can be smaller than that in the case where these are stored for each chord type.

The first address storage unit ADRS stores the start address of each normal pattern described above, and the second address storage unit ADRS2 stores the start address of each variation pattern described above. The start address of each normal pattern can be selected and read according to the group number (any of 0 to 2) and the pattern number (any of 0 to 3), and the start address of each variation pattern is the variation pattern. The data can be selected and read according to the number (0 or 1).

The first pronunciation number storage unit CHASS stores the pronunciation number data representing the simultaneous pronunciation number (one of 1 to 4) of each normal pattern described above, and the second pronunciation number storage unit CHASS2 includes
Polyphony number of each variation pattern described above (1 ~
No. 4) is stored. The pronunciation data of each normal pattern can be selected and read according to the group number and the pattern number as in the case of the head address of the normal pattern described above, and the pronunciation data of each variation pattern can be read. Similarly to the case of the first address of the above, it can be selected and read according to the variation pattern number.

The upper limit pitch storage unit UPLMT stores a key code (79 as an example, corresponding to the G ₅ note) that is the upper limit of the desired accompaniment tone range. The stored key code is used for controlling the pitch of the key code read from the pattern ROM so that the pitch after the modification falls within the accompaniment tone range.

The tempo clock number storage unit QUANT stores the number of tempo clocks per bar according to the time signature of the rhythm.
If the time signature is 24, 24 is stored. If the time signature is 4, 32 is stored.

FIG. 8 shows a storage example of the pattern storage unit PTN. In this example, as in the case of FIG. 7, the 0th to 3rd normal patterns are stored in the storage unit PTNM, the 0th and 1st normal patterns are stored in the storage unit PTNm, and the 0th and 1st normal patterns are stored in the storage unit PTNS.
The second normal pattern is stored in the storage unit PTNV, and the 0th and first variation patterns are stored in the storage unit PTNV.

In FIG. 8, each rectangle such as T1 and T2 indicates a pitch data group forming a pattern, and each pattern is composed of a maximum of four pitch data groups. As an example, the major 0th normal pattern is composed of three pitch data groups T1 to T3, and if this pattern is used, at most 3
Sounds can be pronounced simultaneously. Similarly, for other patterns, the number of rectangles for each pattern corresponds to the number of sounds that can be sounded simultaneously.

Each pitch data group, as representatively shown for T1, is an array of key codes KC necessary for playing a single note for one measure as the address progresses.

The symbols "A", "B", ... "K" attached to each pattern represent the start address of the corresponding pattern.

When the patterns are stored in the pattern storage unit PTN as described above, the storage contents of the first and second address storage units ADRS and ADRS2 and the first and second pronunciation number storage units CHASS
The storage contents of CHASS2 and CHASS2 are as follows. Here, for ADRS and CHASS, group number GNO
And the address specified by the pattern number PNO (GNO, P
NO), and for ADRS2 and CHASS2, the data specified by the variation pattern number VNO is expressed as (VNO).

ADRS (0,0) = A (0,1) = B (0,2) = C (0,3) = D (1,0) = E (1,1) = F (2,0) = G (2,1) = H (2,2) = I ADRS2 (0) = J (1) = K CHASS (0,0) = 3 (0,1) = 2 (0,2) = 2 (0, 3) = 3 (1,0) = 2 (1,1) = 1 (2,0) = 1 (2,1) = 1 (2,2) = 3 CHASS2 (0) = 2 (1) = 4 FIG. 9 shows an example of an accompaniment pattern. In this example, the highest pitch T1, the second highest pitch T2, the third highest pitch T3, and the lowest pitch T4 are four.
It includes a group.

FIG. 10 shows the accompaniment pattern of FIG. 9 for the pitch data groups T1 to T4.
The data format when stored in the pattern storage unit PTN is shown, and for convenience sake, unlike FIG. 8, the arrangement of the key code KC is shown for address progression in the horizontal direction.
When the key code value is 0, it means that the key is off (non-sounding).

Address progression is 0 for A, where A is the start address.
It is shown by adding 1, 2, ... In the case of 4-beat, there are 32 addresses (for example, A + 0 to A + 31 in T1) for one pitch data group, and the above-mentioned storage unit QUANT
By multiplying the value (32) of each by 0, 1, 2, 3 and adding to A, the start address A + 0, A + 32, A + 64, A + 96 of each of the pitch data groups T1, T2, T3, T4 can be determined. . Then, when a count value of a tempo clock counter, which will be described later, is added to each head address, if there is a key code to be sounded at that count value, up to four sounds can be sounded simultaneously. For example, if the count value of the tempo clock counter is
If it is 31, it will address A + 31 for T1, A + 63 for T2, A + 95 for T3, and A + 127 for T4, thereby giving the key code value.
Four sounds corresponding to 67, 62, 59 and 48 can be played simultaneously.

In the case of 3 beats, the number of addresses for one pitch data group is 24 and the stored value of the storage unit QUANT is 24, but the addressing method is the same as in the case of 4 beats described above.

Register Group 26 Of the registers included in the register group 26, those relevant to the implementation of the present invention are listed below.

(1) Run flag RUN ... This is a 1-bit register, and 1 indicates that the rhythm is running, and 0 indicates that the rhythm is stopped.

(2) Clock counter CLK ... This is the clock generator 1
It counts the clock signal from 8 and takes a count value of 0 to 95 in one bar, and it becomes 0 at the timing of 96.
Is reset to.

(3) Tempo clock counter TCLK ... This takes a different count value depending on the time signature. In the case of three time signatures, the count value is incremented by 1 for every 4 pulses of the clock signal and is 0 within 1 bar. It takes a count value of 23, is reset to 0 at the timing of 24, and in the case of 4 beats, the count value increases by 1 every 3 pulses of the clock signal and takes a count value of 0 to 31 within one bar. , 32 is reset at 0.

(4) Bar counter BAR ... This has a counter CLK value of 96.
The count value is incremented by 1 at every timing (at the end of the bar) and takes a count value of 0 to 7 within 8 measures, and is reset to 0 at the timing at which the count value becomes 8 (at the end of the bar).

(5) Key code buffer register KEYBUF _{0 to} KEYBUF ₃ ...
These registers respectively correspond to the four tone generation channels of the TG28 and store the key code corresponding to the key pressed by the accompaniment keyboard.

(6) Chord buffer register CHDBUF ... This is an 8-bit register for storing the chord data read from the chord detection table CHDTBL based on the key depression state on the accompaniment keyboard. The chord type data and the root note data are stored in the lower 4 bits.

(7) Chord register for pronunciation CHORD ... This is the register CHD
It receives chord data from BUF and has the same structure as CHDBUF. When the rhythm is stopped, TG28 registers CH
The generation of chords is controlled according to the chord data of ORD.

(8) Old chord register OLDCHD ... This is the register CHORD
It receives chord data that has already been pronounced from CHOR.
Same as D.

(9) Old root register OLDRT ... This is the register OLDCHD
Is to receive root sound data of lower 4 bits.

(10) Root note register ROOT ... This receives the root note data of the lower 4 bits from the register CHORD.

(11) Chord type register TYPE ... This is the register CHORD
From the upper 4 bits of the chord type data.

(12) Group number register GRP ... This is the group number detected using the group number table GRPTBL.
GNO (one of 0 to 2) is set.

(13) Seventh pattern number register SEVPT ... This is the one in which the seventh pattern number (0 to 2) is set.

(14) Pattern key code register PTNKC ... This stores the key code read from the pattern storage unit PTN.

(15) Semitone register RKCNT ... This is the pattern storage P
Halftone data representing the number of halftones corresponding to the note name of the key code read from the TN is stored.

(16) Sequence register PATSEQ ... This stores the sequence data read from the sequence storage unit SEQ.

(17) Pattern number register PTNO ... This is a register
The pattern number PNO (one of 0 to 3) read from PATSEQ is set.

(18) First address register PTADRS ... This is for setting the start address read from the first address storage unit ADRS or the second address storage unit ADRS2.

(19) Second address register PTADRS2 ... This is for setting the read address of the pattern storage unit PTN determined according to the data of the register PTADRS, the counter TCLK, the storage unit QUANT and the like.

(20) Correction data register ADDKC ... This stores the correction data read from the pitch correction table PMTBL.

(21) Number-of-sounds register CH ... This is the first number-of-sounds storage unit.
The pronunciation data read from the CHASS or the second pronunciation memory CHASS2 is stored.

(22) Chord comparison register CMPCHD ... This is an 8-bit register, and the upper 4-bit part is the register OL.
Old chord type data from DCHD is stored, and pitch difference data representing the root pitch difference calculated based on the root data of registers ROOT and OLDRT is stored in the lower 4 bits. .

(23) Sound generation key code register KEYCOD ... This is for storing sound generation key codes created based on the data of the registers PTNKC, ADDKC and ROOT. During rhythm running, the TG28 controls accompaniment sound generation in accordance with the key code of the register KEYCOD.

(24) Old key code registers OLDKEY _{0 to} OLDKEY ₃ ... These registers correspond to the four tone generation channels of the TG28, respectively, and receive a key code from the register KEYCOD and temporarily store it.

In addition, in the flow chart described later, a calculation formula such as a.MOD.b is shown, but this shows the remainder obtained by dividing a by b (integer operation) or the obtained remainder. .

Main Routine (Fig. 11) Fig. 11 shows the processing flow of the main routine.
At step 40, initialization processing is performed in response to power-on or the like. For example, 0 is set in the flag RUN.

Next, at step 42, it is judged if the rhythm start / stop switch (SW) is on. If this determination result is affirmative (Y), the process proceeds to step 44.

In step 44, the value obtained by subtracting the value of the flag RUN from 1 is set in RUN. Therefore, when RUN is 0, RUN = 1 (rhythm running), and when RUN is 1, RUN = 0 (rhythm stop). After this, step 4
Go to 6.

In step 46, it is determined whether RUN is 1. If the result of this determination is affirmative (Y), the routine proceeds to step 48, where the initial set processing relating to autorhythm and auto chord is performed. That is, 0 is set in each of the counters CLK, TCLK, and BAR, 0 is also set in each of the registrars OLDKEY _{0 to} OLDKEY ₃ , and the register OLDCHD is FF data in hexadecimal notation (data in which all bits are 1) Set.

When the process of step 48 is completed, the process proceeds to step 50.
Also, if the determination result of step 42 or 46 is negative (N), the process proceeds to step 50.

In step 50, it is determined whether or not there is a key-on event on the accompaniment keyboard. If the result of this determination is affirmative (Y), the routine proceeds to step 52, where a key depression processing subroutine as described later with reference to FIG. 12 is executed.

When the process of step 52 is completed or when the determination result of step 50 is negative (N), the process proceeds to step 54,
It is determined whether there is a key-off event on the accompaniment keyboard. If the result of this judgment is affirmative (Y), the process moves to step 56,
When a subroutine of key release processing as described later with reference to FIG. 13 is executed, when the processing of step 56 is completed or when the determination result of step 54 is negative (N), the routine proceeds to step 58,
Other processes (for example, tone color selection, rhythm selection, etc.)
Do.

When the process of step 58 is completed, the process returns to step 42,
Repeat the above process.

Key-pressing process subroutine (FIG. 12) In the key-pressing process subroutine of FIG. 12, step 60
Then, it is judged whether or not the registers KEYBUF _{0 to} KEYBUF ₃ are empty.
If the result of this determination is negative (N), it is not possible to retrieve the key depression data, and the routine returns to the routine of FIG.

If the determination result of step 60 is affirmative (Y),
In step 62, the key code corresponding to the key with the key-on event is written in the empty KEYBUF. And step 64
Move on to.

In step 64, the chord data corresponding to the contents of KEYBUF _{0 to} KEYBUF ₃ is read from the chord detection table CHDTBL and registered.
Put it in CHDBUF. Then, the process proceeds to step 66.

In step 66, it is determined whether RUN is 1. If the result of this determination is negative (N), the routine proceeds to step 68, where the chord data of CHDBUF is placed in the register CHORD. And step 7
Go to 0.

In step 70, the TG 28 is controlled in accordance with the chord data of CHORD to generate the chord corresponding to the depressed state. After that, the process returns to the routine of FIG.

If the determination result of step 66 is affirmative (Y), the routine returns to the routine of FIG. In this case RU
Since N = 1 (rhythm running), auto rhythm and auto chord performance are performed according to the clock interrupt routine described later with reference to FIG.

Key release processing subroutine (Fig. 13) In the key release processing subroutine of Fig. 13, step 80
Then, it is determined whether or not there is the same key code as the key (key release) having the key-off event in the registers KEYBUF _{0 to} KEYBUF ₃ .
If the determination result is negative (N), it means that there is a key-off event unrelated to the sound being sounded, and the process returns to the routine of FIG.

If the determination result of step 80 is affirmative (Y),
In step 82, 0 is set in KEYBUF having the same key code. Then, the process proceeds to step 84.

In step 84, it is determined whether the flag RUN is 1. If the determination result is affirmative (Y), the routine returns to the routine shown in FIG. 11 as in the case of step 66 described above.

If the determination result of step 84 is negative (N),
Moving to step 86, the chord data is read from the chord detection table CHDTBL according to the contents of KEYBUF _{0 to} KEYBUF _{3 and} is stored in the register CHDBUF. Then, the process proceeds to step 88.

In step 88, the chord data of CHDBUF is stored in the register CHORD.
Put in. Then, the process proceeds to step 90, and the TG 28 is controlled according to the chord data of CHORD so that the chord corresponding to the key-depressed state after key release is sounded. As a result, if the key corresponding to one of the four tones is released, for example, the remaining three tones will continue to be produced.

When the processing of step 90 is completed, the routine returns to the routine of FIG.

Clock Interrupt Routine (FIG. 14) FIG. 14 shows the processing flow of the clock interrupt routine, which is started by the clock generator 18 each time a clock pulse is generated.

First, in step 100, it is determined whether the flag RUN is 1. If the determination result is negative (N), the process returns to the routine of FIG.

If the determination result of step 100 is affirmative (Y), the process proceeds to step 102. In this step 102, It is determined whether or not the calculation result by the calculation formula In this calculation formula, CLK represents the count value of the counter CLK, and QUANT represents the storage value of the storage unit QUANT.

In the case of 3 beats, since QUANT = 24, 96 / QUANT becomes 4, and the determination result in step 102 becomes affirmative (Y) every 4 clocks. In the case of 4 beats, QUANT = 32, so 96 / QUANT becomes 3, and the determination result of step 102 becomes affirmative (Y) every 3 clocks. If the determination result of step 102 is negative (N), the routine returns to the routine of FIG. 11, but if it is affirmative (Y), step 104.
Move on to.

In step 104, the selected rhythm type and counter TCL
Rhythm generation processing is performed according to the value of K. That is,
The rhythm pattern corresponding to the selected rhythm type is referred to, and if there is a rhythm sound to be sounded at the timing corresponding to the count value of TCLK, it is sounded. Then, the process proceeds to step 106.

In step 106, an accompaniment pronunciation subroutine as described later with reference to FIG. 15 is executed. This subroutine
This is for automatically generating chords using the accompaniment pattern corresponding to the selected rhythm type.

Next, at step 108, the value of TCLK is incremented by 1, and then the routine proceeds to step 110, where it is determined whether the calculation result by the calculation formula TCLK.MOD.8 is 0. In this formula, TCLK is the counter
Indicates the count value of TCLK. When the calculation result is 0, it means that it is the beginning of a beat.

If the determination result of step 110 is affirmative (Y) (beginning of beat), the process proceeds to step 112, and the register CHO
Put the chord data of RD into register OLDCHD. Then, the process proceeds to step 114, where the chord data in the register CHDBUF is set to CHORD.
Put in. Therefore, in the case of auto chord, even if new chord data is set in CHDBUF due to chord change in the middle of a certain beat, the new chord data is set in CHORD at the timing of the beginning of the next beat. Therefore, the accompaniment based on the new chord data can be performed from this timing.

When the process of step 114 is completed or when the determination result of step 110 is negative (N), the process proceeds to step 116 and the value of the counter CLK is incremented by 1. Then, the process proceeds to step 118.

In step 118, it is determined whether the CLK value is 96 (end of bar). If the determination result is negative (N), the process returns to the routine of FIG.

If the determination result in step 118 is affirmative (Y), the process proceeds to step 120 and the counters CLK and TCLK are set to 0. Then, in step 122, the value of the counter BAR is set to 1
After the value is up, the process proceeds to step 124, and it is determined whether the BAR value is 8. If the determination result is negative (N), the process returns to the routine of FIG.

If the determination result of step 124 is affirmative (Y), the process proceeds to step 126 and BAR is set to 0. Thus, when 0 is set to BAR at the end of 8 measures, it means that the automatic accompaniment for 8 measures has ended, and after that, the automatic accompaniment for 8 measures is repeated again with the same content as the previous time.

When the processing of step 126 is completed, the routine returns to the routine of FIG.

Accompaniment pronunciation subroutine (Fig. 15) Step 13 in the accompaniment pronunciation subroutine of Fig. 15
At 0, upper 4-bit data (chord type data) of the register CHORD is set in the register TYPE. Then, the process proceeds to step 132.

In step 132, it is determined whether the value of TYPE is 6 (seventh). If this determination is negative (N), step
Move on to 134.

In step 134, the group number table GRPTBL to T
The group number GNO (chord type group) corresponding to the YPE value (chord type) is detected and set in the register GRP. Then, the process proceeds to step 136.

In step 136, the sequence data corresponding to the group number of the GRP is read from the storage unit SEQ and stored in the register PATSEQ.
Put in. Then, the process proceeds to step 138.

In step 138, the pattern number PNO of the bar corresponding to the value of the counter BAR is read from PATSEQ and placed in the register PTNO. For example, in step 136, the storage unit SEQ in FIG.
When (0) major sequence data is selected and the BAR value is 0, PNO = 0 in the first measure is PT.
Set to NO.

Next, at step 140, the start address of the normal pattern designated by the value of GRP (group number) and the value of PTNO (pattern number) is read from the storage unit ADRS and is stored in the register PTADRS. For example, when the sequence data of SEQ (0) is selected as described above, the storage unit PT
Assuming that pattern data is stored in M as described above with reference to FIG. 8, if GRP = 0 and PTNO = 0, the start address A of the 0th normal pattern in the storage unit PTNM is stored.
Is set in PTADRS. After this, the process proceeds to step 142.

In step 142, as in step 140 above, the GRP
, And the pronunciation data of the normal pattern designated by PTNO are read and placed in the register CH. And step 144
Move on to.

In step 144, the pattern reading sub routine is executed as described later with reference to FIG. This subroutine is for reading out the key code which constitutes the accompaniment pattern and controlling the accompaniment sound generation. After step 144, the routine returns to the routine shown in FIG.

By the way, when the result of the determination at step 132 is affirmative (Y), the routine proceeds to step 145, where a seventh processing subroutine is executed as described later with reference to FIG. This subroutine is to set the pattern number for Seventh of 0 to 2 in the register SEVPT according to the mode of transition to Seventh. After step 145, the process moves to step 146.

In step 146, it is determined whether the value of SEVPT is 0 (a normal pattern for Seventh). This judgment result is positive (Y)
If so, the process proceeds to step 134, and the subsequent processes are executed in the same manner as described above.

If the determination result of step 146 is negative (N), the process proceeds to step 147. In this step 147, SEVPT
The leading address of the variation pattern designated by subtracting 1 from the value of (variation pattern number) is read from the storage unit ADRS2 and placed in PTADRS. Then, the process proceeds to step 148.

In step 148, as in step 147 above, SEVP
The pronunciation data of the variation pattern designated by T-1 is read from the storage unit CHASS2 and placed in CH.

After this, similarly to the above, the process of step 144 is performed and the process returns to the routine of FIG.

Seventh processing sub routine (FIG. 16) In the seventh processing subroutine of FIG. 16, in step 150, the lower 4 bits (root note data) of the register CHORD is placed in the register ROOT. Then, the process proceeds to step 152.

At step 152, the lower 4 bits (root note data) of the register OLDCHD are put into the register OLDRT. Then, the process proceeds to step 154.

In step 154, the calculation result by the calculation formula (ROOT-OLDRT + 12) .MOD.12 is put in the lower 4 bits of the register CMPCHD. In this formula, ROOT is the register RO
The value of OT and the value of OLDRT represent the value of the register OLDRT, and the addition of 12 to (ROOT-OLDRT) is to prevent the difference from becoming negative. According to this formula, it is possible to find the root pitch difference (the number of semitones from 0 to 11) between the 7th chord and the chord immediately before it.
Pitch difference data representing the root pitch difference is stored in the bit portion.

Next, in step 156, the upper 4 bits of OLDCHD (old chord type data) are set in the upper 4 bits of CMPCHD. Then, the process proceeds to step 158.

In step 158, the control variable i is set to 0. Then, the process proceeds to step 160, where i in the chord progression determination table SEQREF is
It is determined whether the determination reference data of the second storage area SEQREF _{i and} the data of CMPCHD match. This judgment result is negative (N)
If so, go to step 162.

In step 162, the value of i is incremented by 1. Then, the process proceeds to step 164 and it is determined whether i is 12 or more. This step 1
When i = 0 is set in step 58 and the first time step 164 is reached, since i = 1, the determination result is negative (N).
Return to step 160. In this way, the storage area SEQREF ₀ ~
If no match is obtained even when the contents of SEQREF ₁₁ are sequentially compared with the contents of CMPCHD, the determination result of step 164 becomes affirmative (Y), and the process proceeds to step 166.

In step 166, 0 is set in the register SEVPT. Then, the process returns to the routine of FIG. In this case, in the routine of FIG. 15, the processing of steps 134 to 144 is performed, and the accompaniment sound is generated according to the normal pattern for Seventh.

In the above sequential comparison, when the determination result of step 160 is affirmative (Y) (a match is obtained), the process proceeds to step 168.

In step 168, it is determined whether i is greater than 7 (whether a transitional accompaniment pattern should be selected). If the determination result is negative (N), the process proceeds to step 170 and SEVPT is set to 1. Then, the process returns to the routine of FIG.

If the determination result of step 168 is affirmative (Y), the process proceeds to step 172, and SEVPT is set to 2. Then, the process returns to the routine of FIG.

When the routine returns to the routine shown in FIG. 16 through step 168 or 172, step 14 in the routine shown in FIG.
The processes of 7, 148 and 144 are performed, and the accompaniment sound is generated according to the variation pattern of the variation pattern number 0 or 1.

Pattern Reading Subroutine (FIG. 17) In the pattern reading subroutine of FIG. 17, in step 180, the control variable i is set to zero. Then, the process proceeds to step 182.

In step 182, the calculation result by the calculation formula PTADRS + TCLK + i × QUANT is put into the register PTADRS2. In this formula, PTADRS represents the value of the register PTADRS, TCLK represents the value of the counter TCLK, and QUANT represents the storage value (24 or 32) of the storage unit QUANT. According to this calculation formula, the read address of the pattern memory PTN can be determined, and the PTA
The determined read address is set in DRS2. For example, when step 182 first comes after step 180, i = 0. Therefore, if PTADRS is A shown in FIG. 9 and TCLK = 0, PTADRS2 has the start address of T1 in FIG. (A + 0) is set. After this step
Move to 184.

In step 184, the data of the address designated by PTADRS2 is read from the storage unit PTN and is stored in the register PTNKC.
Then, the process proceeds to step 186.

In step 186, it is determined whether the value of PTNKC is 0 (non-sounding). If the determination result is negative (N), step 18
Move to 8 and put the calculation result by the calculation formula PTNKC.MOD.12 into the register RKCNT. In this calculation formula, PTNKC represents the key code value of the register PTNKC. According to this formula, the number of semitones (0 to 1) corresponding to the note name of the PTNKC key code
1)), and the root note C can be found in RKCNT.
Halftone data representing the number of halftones from is stored. After this, the process proceeds to step 190.

In step 190, the pitch correction table PMTBL is used to register
The correction data corresponding to the contents of TYPE and RKCNT is read and placed in the register ADDKC. Then, the process proceeds to step 192.

At step 192, the calculation result by the calculation formula PTNKC + ADDKC + ROOT is put into the register KEYCOD. In this formula, PTNKC, ADDKC and ROOT are the registers PTNK
The values of C, ADDKC and ROOT are shown. According to this calculation formula, the key code of the putter stored with the root note C as a reference is changed in pitch according to the root note of the chord specified by the accompaniment keyboard, and at the time of changing the pitch, there is no dissonance. The pitch can be corrected according to the correction data so that it does not become a sound. For example, if you specify a chord A _m in the accompaniment keyboard, TYPE = 3, ROOT
= 9. In this case, assuming that PTNKC = 40 (E ₂ ), RKCNT becomes 4 and ADDKC becomes −1. Therefore, the KEYCOD, 40-1 + 9 = 48 ( C 3) is set. After step 192, the process moves to step 194.

In step 194, it is determined whether the value of KEYCOD is larger than the value of the storage unit UPLMT (whether it exceeds the upper limit pitch). If the result of this judgment is affirmative (Y), then the procedure proceeds to step 196, where the value obtained by subtracting 12 from the KEYCOD value (down one octave) is KEYC.
Set to OD. Then, the process returns to step 194 and the same determination as above is performed again. Such processing is performed until the value of KEYCOD becomes equal to or lower than the value of UPLMT, and the value of KEYCOD finally corresponds to a pitch equal to or lower than the upper limit pitch.

When the determination result of step 194 is originally negative (N) or when the result of the process of step 196 is negative (N), the process proceeds to step 198.

In step 198, it is determined whether the value of KEYCOD matches the value of the i-th old key code register OLDKEY _i (previous key code value). If the result of this judgment is negative (N), the key code of KEYCOD is set to OLDKEY _i in step 200, then the flow proceeds to step 202, and TG2 is set according to the key code of KEYCOD.
Controls the tone generation of the 8th channel i (channel 0 at the beginning). As a result, if the key code value 48 is set in the KEYCOD as in the above-mentioned example, the C ₃ tone corresponding thereto is sounded. In addition, when a next sound is sounded while a certain sound is being sounded, the previous sound is rapidly attenuated and then the subsequent sound is sounded.

As described above, the determination result of step 186 is negative (N).
When the determination result of step 186 is affirmative (Y), the processing flow is as follows. That is, in this case, the value of PTNKC is 0.
Since it is (non-pronunciation), the process proceeds to step 204 without going through the processes of steps 188 to 196.

At step 204, the value (0) of PTNKC is put into KECOD.
Then, in step 198, the KEYCOD value (0) is OLDKE.
It is determined whether or not it matches the value of Y _i . This judgment result is negative (N)
If so, move to step 200 and set KEYCOD value (0) to OLD
Put it in KEY _i .

After that, in step 202, T is changed according to the value (0) of KEYCOD.
The i-th channel of G28 (the 0th channel at the beginning) is controlled to stop sounding.

Whether the determination result in step 186 is negative or positive, the process proceeds to step 206 after step 202. If the determination result of step 198 is affirmative (Y), it is not necessary to start or stop the sound generation, and thus the process proceeds to step 206 without passing through steps 200 and 202.

In step 206, the value of i is incremented by 1. Then, the process proceeds to step 208, and it is determined whether i is smaller than the value of the register CH. The CH stores the pronunciation data indicating the number of notes that can be simultaneously pronounced for the selected accompaniment pattern. If the value of this data is 1, when step 208 first comes to step 208, Is 1, the determination result is negative (N), and the process returns to the routine of FIG. This is the case, for example, where the pattern is composed of a single pitch data group, such as the pattern shown in FIG.

When the pattern is composed of a plurality of pitch data groups as shown in, for example, T1 to T3 of FIG. 8, the determination result of step 208 becomes affirmative (Y), and the process returns to step 182. Then, in step 208, the above processing is repeated until i = CH, and when i = CH, the process returns to the routine of FIG. As a result, for example, as described above with reference to FIG. 10, a plurality of tones (up to 4 tones) can be sounded simultaneously.

Modifications The present invention is not limited to the above-described embodiments, but can be implemented in various modified forms. For example,
The present invention can be applied to an auto base and the like.

As described above, according to the present invention, the sequence data and the plurality of pattern data are stored for each chord type group, and the pitch of the accompaniment sound signal is corrected according to the pitch correction data. Therefore, it is possible to further reduce the storage capacity and obtain the effect of preventing the generation of dissonance.

[Brief description of drawings]

FIG. 1 is a block diagram showing a circuit configuration of an electronic musical instrument having an automatic accompaniment apparatus according to an embodiment of the present invention, FIG. 2 is a diagram showing key code values for each pitch, and FIG. The figure which shows the memory content of the memory | storage part regarding the root note C in the chord detection table CHDTBL, FIG. 4 is a figure which shows the memory content of the group number table GRPTBL, and FIG. 5 is the figure which shows the memory content of the pitch correction table PMTBL. FIG. 6 is a diagram showing stored contents of a chord progression determination table SEQREF. FIG. 7 is a diagram showing stored contents of a pattern ROM. FIG. 8 is a diagram showing stored examples of a pattern storage unit PTN. FIG. 10 is a staff diagram showing an example of an accompaniment pattern, FIG. 10 is a diagram showing a storage data format of the accompaniment pattern of FIG. 9, FIG. 11 is a flowchart showing a main routine, and FIG. A flowchart showing a processing subroutine, FIG. 13 is a flow chart showing a key release processing subroutine, FIG. 14 is a flow chart showing a clock interrupt routine, FIG. 15 is a flow chart showing an accompaniment sounding subroutine, and FIG. 16 is a seventh processing subroutine. Flowchart, FIG. 17 is a flowchart showing a pattern reading subroutine. 10 ... bus, 14 ... accompaniment keyboard circuit, 16 ... control switch group,
18 ... Clock generator, 20 ... Central processing unit, 22 ... Program memory, 24 ... Accompaniment data memory, 26 ... Register group, 28
… Tone generator, 30… Sound system.

Claims (1)

[Claims] 1. A keyboard for accompaniment, (b) a detecting means for detecting a chord type and a chord root note on the keyboard based on a key depression state, and (c) a plurality of predetermined chord types. The accompaniment data corresponding to each chord type group is a first storage unit that stores accompaniment data for each chord type group of the group. A plurality of M pattern data represented by an array of data and first to Nth patterns for designating these pattern data by appropriately combining them.
(N is an integer larger than M), and (d) determination means for determining which of the plurality of chord type groups the chord type detected by the detection means belongs to, e) selecting means for selecting in the first storage means accompaniment data corresponding to the chord type group determined to belong to this determining means; and (g) chord types belonging to the plurality of chord type groups. Second storage means for storing pitch correction data for preventing dissonance in correspondence with a predetermined chord type and a predetermined note name; and (h) sequentially according to sequence data in the accompaniment data selected by the selection means. By accompaniment sound generating means for generating an accompaniment sound signal having a pitch determined by the pitch data in the pattern data and the chord root detected by the detecting means by reading the pattern data in the accompaniment data. If the predetermined chord type is detected by the detection means and the pitch data of the predetermined pitch name is read from the first storage means, the pitch correction data of the second storage means is read. According to the automatic accompaniment device for electronic musical instruments, the pitch of the accompaniment sound signal is corrected so as not to cause dissonance.