JPH04240893A - Automatic accompaniment instrument - Google Patents

Abstract

PURPOSE:To automatically generate song data while imaging a flow of a large accompaniment and to eliminate input misoperation by classifying accompaniment patterns with specific indexes and storing the data, and automatically generating the song data with the indexes. CONSTITUTION:A pattern memory 6 is stored with pattern data showing the accompaniment patterns. The pattern data are classified into a hierarchic structure, e.g. a 1st layer of rock, pops, etc., a 2nd layer of introductions, endings, etc., and a 3rd layer of normal rhythm, filling-in, etc. When the pattern memory 6 is accessed, an index is specified. In song play mode, automatic performance is carried out according to the song data.

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic accompaniment device, and more particularly to an automatic accompaniment device that can easily generate song data for performing automatic accompaniment. [0002] Conventionally, a so-called pattern memory stores a plurality of different accompaniment patterns representing the accompaniment content of a certain section (for example, one bar), and a pattern for specifying this accompaniment pattern in a song memory. A plurality of pieces of specification information are stored, and the pattern specification information of this song memory is sequentially read out at a predetermined tempo, and an accompaniment pattern is read out based on the read pattern specification information, and various rhythm tones and bass tones are played according to the accompaniment pattern. Automatic accompaniment devices that produce sounds such as For example, Japanese Patent Publication No. 1-39117 discloses an automatic accompaniment device that stores pattern designation information that designates an accompaniment pattern for each bar. Also, when repeating the same accompaniment pattern in multiple consecutive measures,
Some song memories store data indicating repeats. FIG. 15 shows an example of song data stored in a song memory in a conventional automatic accompaniment device. The song data includes pattern specification information specifying an accompaniment pattern and data indicating repeat, and is stored in continuous areas PART001, PART002, PART003 of the song memory.
,... are stored sequentially. Here, each area PART0
01, PART002, . . . each correspond to one bar. [0005] For automatic accompaniment, first, the first area PART001 of the song data is read out. Area PART
Since information specifying pattern 00 (for example, the address of the pattern in the pattern memory) is written in 001, this accompaniment pattern 00 is read out from the pattern memory, and automatic accompaniment is performed with the accompaniment pattern according to this pattern 00. Ru. Next, an area PART002 stores data indicating the start of repeat, an area PART003 stores a repeat pattern, and an area PART004 stores data indicating the end of the repeat range and the number of repeats. Therefore, accompaniment pattern 01 is automatically performed repeatedly four times. [0006] When storing the number of repeats as in the area PART004, the data at the end of the repeat range of this area PART004 is read out (at this point, the accompaniment pattern to be repeated has already been read out once). ) pattern to be repeated again after
Since the number of times the song returns to is stored, the number of times the automatic performance is actually performed is one more than the number of times that is stored. therefore,
In the example of FIG. 15, pattern 01 is repeated four times. Conventionally, when generating song data as described above, a desired one is selected from among a plurality of accompaniment patterns (for example, about 100 types) for each predetermined section such as one measure, and the song data is sequentially created. It was necessary to set each area of memory. [0008] In the conventional method of setting song data, accompaniment patterns are selected one by one for each predetermined section, so the problem is that it is difficult to understand the overall flow of the accompaniment. was there. Furthermore, since one of many types of accompaniment patterns is selected, the selection process is complicated and input errors are often made. In view of the above-mentioned problems in the conventional example, the present invention makes it possible to set song data for automatic accompaniment by imagining the general flow of the accompaniment, and also allows for easy input operations with fewer erroneous inputs. The purpose is to provide an automatic accompaniment device. [Means for Solving the Problems] In order to achieve this object, the automatic accompaniment device according to the present invention classifies a plurality of different accompaniment patterns representing the accompaniment content of a certain section using a predetermined index. In order to generate desired song data in which a plurality of stored first storage means and the accompaniment pattern itself or pattern designation information specifying the accompaniment pattern are arranged, parameters including an index specifying an accompaniment pattern are input. parameter input means;
The present invention is characterized by comprising song data generation means for generating the song data according to the parameters input by the parameter input means. Further, a second storage means for storing the song data generated by the song data generation means;
It is preferable to include accompaniment means for performing automatic accompaniment based on the song data stored in the second storage means. Preferably, the first storage means stores a plurality of accompaniment patterns in a hierarchical manner using a predetermined index. [0012] Since the accompaniment patterns are classified by a predetermined index, the accompaniment pattern can be specified using this index. Therefore, by inputting parameters including this index using the parameter input means, song data can be automatically generated by the song data generation means. The generated song data is stored in the second storage means and read out by the accompaniment means to perform automatic accompaniment. As a storage method for accompaniment patterns, if accompaniment patterns are stored in hierarchies using predetermined indexes, song data can be generated by sequentially specifying hierarchies. Embodiments [0013] Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 shows a block configuration of an automatic accompaniment apparatus according to an embodiment of the present invention. This automatic accompaniment device includes a central processing unit (CPU) 1 that controls the operation of the entire device, an operation panel 2 equipped with various switches and a display device, etc.
and pad 3. The pad 3 is used to generate a percussion instrument sound when pressed in the normal mode (the mode will be described later), and to input an accompaniment pattern in the pattern edit mode. The automatic accompaniment device further includes a working memory 4, a program memory 5, a pattern memory 6, and a song memory 7. The working memory 4 is used for various work registers, flags, etc. The program memory 5 stores programs executed by the CPU 1 and the like. The pattern memory 6 stores pattern data representing accompaniment patterns. Song memory 7 stores song data. Reference numeral 8 indicates a clock that issues a timer interrupt to the CPU 1 at predetermined time intervals (in this case, 1 msec), 9 indicates a tone generator (TG), 10 indicates a sound system, and 11 indicates a bidirectional bus line. FIG. 2 shows the appearance of the operation panel 2. As shown in FIG. The operation panel 2 includes a numeric keypad 21, a cursor key 22, a display device 23, a pattern switch 24, a song switch 25,
It has a rest switch 26, a make switch 27, a start switch 28, and a stop switch 29. The cursor keys 22 include an upward cursor key 22a, a downward cursor key 22b, a right cursor key 22c, and a left cursor key 22d. Next, modes in the automatic accompaniment apparatus shown in FIGS. 1 and 2 will be explained. This automatic accompaniment device has the following six modes. (1) Normal mode This is a mode for normal performance. By inputting predetermined parameters in advance from the operation panel 2 and pressing the pad 3, a percussion instrument sound of a predetermined tone is produced. (2) Pattern mode In order to create an accompaniment pattern for each predetermined section, for example, one bar, this is a mode in which an index for specifying the accompaniment pattern is set. (3) Pattern Edit Mode This mode is for creating the contents of the accompaniment pattern specified by the index set in the pattern mode. For example, input the timing within one bar to generate a percussion instrument sound, or the timing to generate a rest. (4) Song mode In this mode, in order to create song data, parameters including an index specifying an accompaniment pattern constituting the song data are input. (5) Song Make Mode According to the index entered in the song mode above,
This mode automatically generates song data. (6) Song Play Mode This is a mode in which automatic accompaniment is performed based on stored song data. FIG. 3 is a state transition diagram of the automatic accompaniment device of this embodiment. When the pattern switch 24 is pressed in the normal mode, the mode shifts to the pattern mode. When the pattern switch 24 is pressed again in the pattern mode, the mode returns to the normal mode. When the start switch 28 is pressed in the pattern mode, the mode shifts to the pattern edit mode. When the stop switch 29 is pressed in the pattern edit mode, the mode shifts to the normal mode. The pattern mode has four submodes. In each submode, set the accompaniment pattern index. Pattern switch 2 from normal mode
When you press 4 to shift to the pattern mode, the mode becomes the style setting submode. In the style setting submode, the style of accompaniment pattern data to be created is set. A style is an index indicating a musical type such as rock or pop. When the downward cursor key 22b is pressed in the style setting submode, the mode shifts to the intro solo ending setting submode, and when the upward cursor key 22a is depressed, the mode shifts to the quantize setting submode. In the intro solo ending setting submode, an index related to the intro solo ending is set. That is, it is set whether the accompaniment pattern data to be created from now on is an introduction (introduction) pattern, a solo (main body part following the introduction) pattern, or an ending (final part) pattern. In the intro solo ending setting submode, pressing the upward cursor key 22a shifts to the style setting submode, and pressing the downward cursor key 22b shifts to the normal fill-in setting submode. In the normal fill-in setting submode,
Set the index related to normal fill-in. That is, it is set whether the accompaniment pattern data to be created from now on is a normal pattern or a fill-in pattern. The normal pattern is a normal accompaniment pattern, and the fill-in pattern is an accompaniment pattern inserted at an appropriate position in place of the normal accompaniment pattern. In the normal fill-in setting submode, pressing the upward cursor key 22a shifts to the intro solo ending setting submode, and pressing the downward cursor key 22b shifts to the quantize setting submode. In the quantization setting submode, quantization of accompaniment pattern data to be created from now on is set. Quantize is a value indicating how much resolution this accompaniment pattern is divided into. Percussion instrument sounds to be generated can be set in the pattern edit mode using sections divided according to the quantization value as units. In the quantize setting submode, pressing the upward cursor key 22a shifts to the normal fill-in setting submode, and pressing the downward cursor key 22b shifts to the style setting submode. [0028] In normal mode, song switch 2
Press 5 to switch to song mode. When the song switch 25 is pressed again in song mode, the mode returns to normal mode. Make switch 2 in song mode
Pressing 7 moves to song make mode. When the make switch 27 is pressed again in the song make mode, the mode returns to the song mode. When the start switch 28 is pressed in the song mode, the mode shifts to the song play mode. In song make mode, press start switch 2
Pressing 8 shifts to song play mode. When the stop switch 29 is pressed in the song play mode, the mode shifts to the normal mode. Song mode has seven submodes. In each submode, set parameters for automatically generating song data. This parameter contains an index that identifies the pattern. When the song switch 25 is pressed from the normal mode to the song mode, the song number setting submode is entered. In the song number setting submode, the song number of song data to be created is set. The song number is a number that identifies song data. Down cursor key 2 in song number setting submode
When 2b is pressed, the mode shifts to the style setting submode, and when the upward cursor key 22a is depressed, the mode shifts to the fill-in setting submode. In the style setting submode, the style of the song data to be created is set. When the downward cursor key 22b is pressed in the style setting submode, the mode shifts to the tempo setting submode, and when the upward cursor key 22a is depressed, the mode shifts to the song number setting submode. [0032] Below, other sub-modes of the song mode include a tempo setting sub-mode, a sub-mode for setting the number of measures for the entire song, a sub-mode for setting the number of solo starting measures, a sub-mode for setting the number of solo ending measures, and a sub-mode for setting the number of measures in the fill-in. Although there is a mode, it is omitted in FIG. Transition between these submodes is performed by pressing the upward cursor key 22a or the downward cursor key 22b. These submodes are for setting various parameters regarding song data to be created. In the tempo setting submode, the tempo of the song is set. The number of measures for the entire song setting submode sets the number of measures for the entire song. In the solo start measure number setting submode, you can set the measure at which the solo section starts. In the solo ending measure number setting submode, you can set the measure at which the solo section ends. In the fill-in measure setting sub-mode, you can set at which measure (or more than one measure) the fill-in will be inserted. Next, operations and screen displays in the pattern mode and pattern edit mode will be explained with reference to FIG. As described above, when the pattern switch 24 is pressed in the normal mode, the style setting submode of the pattern mode is entered. 41 is an example of a display screen in the style setting submode. A message prompting the user to input a style number is displayed on the screen. 46 is a cursor. When a style number is input using the numeric keypad 21, the number is displayed at the position of the cursor 46 and is set as the style number. Next, when the downward cursor key 22b is pressed, the intro solo ending setting submode is entered. 42 is an example of a display screen in the intro solo ending setting submode. A message is displayed on the screen prompting you to select an intro, solo, or ending. Right cursor key 22c or left cursor key 2
By pressing 2d, move the cursor 46 to "
It can be set to ``INTRO,''``SOLO,'' or ``ENDING,'' allowing you to select an intro, solo, or ending. Next, when the downward cursor key 22b is pressed, the normal fill-in setting submode is entered. 43
is an example of a display screen in the normal fill-in setting submode. A message prompting you to select normal or fill-in is displayed on the screen. The selection method is the same as the intro solo ending setting above. Next, when the downward cursor key 22b is pressed, the quantize setting submode is entered. 44 is an example of a display screen in the quantize setting submode. Four selection techniques for quantization are displayed on the screen. "1/16" indicates that a rhythm sound can be set to be produced in units of 16 equal sections within one bar. “1/
Similarly, ``8'', ``1/4'' or ``1/2'' are each 1
This shows that each section of a measure can be divided into 8 equal parts, 4 equal parts, or 2 equal parts, and can be used as a unit. The selection is made by moving the cursor in the same way as the intro solo ending setting above. After setting the data in the pattern mode as described above, pressing the start key enters the pattern edit mode. 45 is an example of a display screen in the pattern edit mode. One horizontal line 47 is one beat (
beat) represents the position. The number of horizontal lines 47 is set according to the set quantization. For example, quantize is 1/
If it is 16, 16 horizontal lines 47 are displayed. Numbered 45
The display example shows when the quantization is 1/4, and the horizontal line 4
Four 7s are displayed. The vertical line 53 indicates the end of the pattern (in this case, the end of the measure), and the black circles 50 displayed above and below the horizontal line 47 indicate that a predetermined rhythm sound will be produced at that timing. The timbre of the rhythm tones generated in the upper stage 48 and the lower stage 49 can be made different. Inputting a pattern, that is, inputting a black circle 50, is performed by positioning the cursor 46 at a desired beat position using the cursor movement keys 22 and pressing the pad. When the black circle 50 is not set at the position of the horizontal line 47, the timing is a rest. A rest is input by positioning the cursor 46 at a desired beat position using the cursor movement keys 22 and pressing the rest switch 26. In the manner described above, the contents of the accompaniment pattern corresponding to the index input in the pattern mode are set. After creating a predetermined accompaniment pattern, song data is created. FIG. 5 shows operation and screen display examples of the song mode, song make mode, and song play mode. As described above, when the song switch 25 is pressed in the normal mode, the song number setting submode of the song mode is entered. 51 is an example of a display screen in the song number setting submode. A message prompting you to input the song number is displayed on the screen. When a song number is input using the numeric keypad 21, the number is displayed at the position of the cursor 46 and is set as the song number. Next, when the downward cursor key 22b is pressed, the style setting submode is entered. 52 is an example of a display screen in the style setting submode. A message prompting the user to enter a style number is displayed on the screen. A style number can be input using the numeric keypad 21. [0045] Similarly, the tempo setting submode,
Parameters for generating song data are input in each of the sub-mode for setting the number of measures for the entire song, the sub-mode for setting solo start measures, the sub-mode for setting solo end measures, and the sub-mode for setting fill-in measures. After inputting the parameters, the make switch 27 is pressed to enter the song make mode, and song data is automatically created. Song data is automatically created based on the parameters input as described above. After creating the song data, when the start switch 28 is pressed, the song play mode is entered and automatic accompaniment starts based on the created song data. FIG. 6 shows the hierarchical structure of pattern data in the automatic accompaniment device of this embodiment. The pattern data is classified by style, which is one of the indexes, in the first hierarchy at the top. The style indicates, for example, rock or pop. In the second layer, songs are classified by index indicating whether they are intro, solo, or ending. Furthermore, as the next third layer,
It is usually classified by an index that indicates whether it is a rhythm or a fill-in. PTN indicates a pattern memory that stores pattern data. PTN (i, j, k) indicates pattern data for one measure. Note that PTN (i, j, k) indicates a storage position in the pattern memory and also indicates data stored at that position. The pattern data is classified in the hierarchical structure described above, and the pattern memory is accessed by specifying its index. That is, the pattern memory PTN (
i, j, k), the first parameter i is a value indicating the style of the first layer, the second parameter j is a value indicating whether the second layer is intro, solo, or ending, and the third parameter k is
specifies a value indicating whether the third layer is a normal rhythm or a fill-in. Here, the value indicating rock as a style is set to "0", the value indicating pop is set to "1", and so on. Further, the value indicating an intro was set as "0", the value indicating a solo was set as "1", and the value indicating an ending was set as "2". Further, a value indicating a normal rhythm was set as "0", and a value indicating a fill-in was set as "1". Therefore, for example, the normal rhythm used for a rock intro is PTN (0, 0, 0), and the fill-in rhythm used for a rock intro is PTN (0, 0, 0).
0,1), and the normal rhythm used for rock solos is PTN.
(0, 1, 0), etc. can be accessed as follows. FIG. 7 shows an accompaniment pattern (pattern data)
Here is an example of arranging them to form one song. A song can be roughly divided into an intro section, a solo section, and an ending section. Each section is constructed by appropriately arranging a normal rhythm and a fill-in rhythm. For example, in the example shown in FIG. 7, the intro section is set to 8 measures, and the first 3 measures are the normal rock intro rhythm PTN (0, 0, 0
) is repeated three times. "x3" written on the upper right of PTN (0, 0, 0) indicates that the same pattern is repeated three times. Next, the rock intro fill-in rhythm PT
N(0,0,1) is played once and then the normal rhythm PT is played again.
N(0,0,0) is repeated four times. [0050] The solo section following the intro section is set to 8 measures, and the first 3 measures are the normal rhythm PT of a rock solo.
N(0,1,0) is repeated three times. Next, the rock solo fill-in rhythm PTN (0, 1, 1) is played once, and the normal rhythm PTN (0, 1, 0) is repeated four times. [0051] The ending section that follows the solo section is set to 8 bars, and is the normal rhythm PTN of a rock ending.
(0, 2, 0) is repeated 8 times. FIG. 8 shows specific contents of song data corresponding to the song having the structure shown in FIG. PART indicates a song memory that stores song data. PART(
SONG, n), the nth song number SONG (n
shall be counted in ascending order starting from 0). It is assumed that PART (SONG, n) indicates the storage location in the song memory and also indicates the data stored at that location. In FIG. 8, element PART(SONG,
0) to PART (SONG, 2) correspond to a portion in which the pattern PTN (0, 0, 0) of the first three bars of the intro is repeated three times. That is, store the repetition start symbol (the image of this repetition start symbol is shown below. Hereinafter, "outside 1" shall indicate this image) in the element PART (SONG, 0), and then PART(SONG,
1) stores the pattern PTN (0, 0, 0) to be repeated (the pattern itself may be stored, or an address specifying the pattern may be stored),
Furthermore, the next element PART (SONG, 2) has a repetition end symbol (the image of this repetition end symbol is shown below. Hereinafter, "outside 2" will refer to this image) and the previous (i.e. PART The number of times to return (to (SONG, 1)) "x2" is stored. [Example 1] [Example 2] The next element PART (SONG, 3) corresponds to the intro fill-in pattern PTN (0, 0, 1), and its pattern (or address) is It is stored. Continue from PART(SONG, 4) to PART(SONG)
G, 6) has the intro pattern PTN (0, 0, 0)
It stores repetition data corresponding to the part where is repeated four times. Thereafter, data corresponding to FIG. 7 is stored in the same manner, and an end code is stored in the last element PART (SONG, 17), thereby completing one song data. Next, the registers used in the automatic accompaniment device of this embodiment described above will be explained. (a) MOD: Mode register indicating mode. As shown in Figure 3, "0" in normal mode, "1" in pattern mode, "2" in pattern edit mode, "3" in song mode, "4" in song make mode, "5" when in song play mode
Take the value of each. (b) T: The time between beats in units of 1/16 beats is 1 msec.
This is a counter for counting down in units. (c) BT: 1m representing the time between beats in units of 1/16 beats
This is a time register that stores a value in seconds. (d) Q: This is a counter that counts up every 1/16 beat within one measure of the accompaniment pattern. (e) BEET: This is a beat register indicating the beat of the pattern. The unit is 1/4 beat, and for example, "3" is set for a song with a 3/4 time signature, and "4" is set for a song with a 4/4 time signature. (f) PATERN: A pattern register that temporarily stores an accompaniment pattern to be played. (g) POINT: A pointer pointing to the next element of the song data element to be played. (h) BAR: A counter that counts the number of bars used when creating song data. (i) BSPT: A normal pattern register that temporarily stores normal patterns. (j) FLPT: A fill-in pattern register that temporarily stores a fill-in pattern. (k) FIN: This is a boundary pointer that points to the boundary position (indicated by the number of measures from the beginning) of the intro section, solo section, and ending section. (l) RPN: A repeat counter that counts down the number of accompaniment pattern repetitions when playing a song. (m) REPEAT: This is a repeat state register indicating the repeat state of the accompaniment pattern. Set to “1” when the repeat start symbol is read,
The next time you read the repetition end symbol and the number of repetitions, it will be ``0''.
is set to (n) SONG: A song number register that stores a song number. (o) STYL: A style number register that stores a style number. (p) TMP: Tempo register that stores the tempo value. Tempo is expressed by the number of quarter notes included in one minute. (q) NBAR: This is a total bar number register that stores the number of bars in the entire song. (r) ST: A solo start measure register that stores the start measure position (number of measures from the beginning) of the solo section. (s) END: This is a solo end measure register that stores the end measure position (number of measures from the beginning) of the solo section. (t) FLL(i): This is a fill-in measure register that stores the measure position where a fill-in pattern is inserted. Since the fill-in pattern may be inserted at multiple positions, it is in an array format (i is 0, 1, . . . ). (u) FN: Fill-in count register indicating the number of fill-ins. (v) PN: Pointer pointing to song memory. [0060] The symbols indicating each register type represent the register itself as well as the contents of the register (
value). Next, the operation of the automatic accompaniment device of this embodiment will be explained with reference to FIGS. 9 to 14. Referring to the main routine of FIG. 9, when the operation of this apparatus starts, first, in step S1, initial settings of each register, etc. (for example, MOD=0) are performed. Next, in step S2, it is determined whether or not there is a pad switch event. If there is a pad switch event, the mode register MOD is set to "2" in step S3.
”, that is, whether the mode is pattern editing mode. When in the pattern edit mode, the beat position of the pattern in the pattern memory selected based on the style, intro solo ending, normal fill-in, quantize, and cursor position information is determined in step S4 (displayed in Figure 4). 45)), the data (black circle 50 in display 45) is written. Next, in step S5, sound generation and muting processing is performed, and the process proceeds to step S6. If the mode MOD is not "2" in step S3, it is not the pattern edit mode, so writing to the pattern memory is not performed, and the process directly advances to step S5. If there is no pad switch event in step S2, the process directly advances to step S6. [0064] In step S6, it is determined whether there is an event related to the switch. If so, the switch routine (FIG. 10) is called in step S7, and the process returns to step S2. If there is no switch event in step S6, the process returns to step S2. Repeat the above process. Referring to the flowchart of FIG. 10, when the switch routine is called, first in step S11 the switch in which the event occurred is determined. In the case of an event of the pattern switch 24, song switch 25, make switch 27, start switch 28, stop switch 29, or the upward or downward cursor keys 22a and 22b, the process advances to step S12. If it is an event on the numeric keypad 21, the process advances to step S18. In the case of an event of the rightward or leftward cursor keys 22c, 22d, the process advances to step S19. If the event is the rest switch 26, the process advances to step S21. The case where the process proceeds to step S12 is when the switch for changing the mode or submode is pressed as explained with reference to FIG. Therefore, in step S12, the mode and display screen are switched, and the mode register MOD is set to a value according to the state. Next, step S
At step 13, it is determined whether the value of the mode register MOD is "4" or not, that is, whether or not it is the song make mode. In the song make mode, the make routine (FIG. 11) is called in step S14, and then the process returns. If it is determined in step S13 that the mode is not the song make mode, it is determined in step S15 whether the value of the mode register MOD is "5" or not, that is, whether or not the mode is the song play mode. When the song play mode is selected in step S15, the time register BT and counter T are set in step S16.
, counter Q, pointer POINT, repeat counter RPN, and repeat status register. Specifically, the time register BT has 60×
Set 103/(TMP×4) by rounding off the decimal point. The value of tempo register TMP is 1 minute (=60×
103msec) indicates the number of quarter notes, so 60×10
3/TMP is the time of one quarter note, that is, the time interval between beats in units of quarter notes expressed in msec. This is further divided by 4 to calculate a value representing the time from one beat to the next beat in units of 16th notes in msec notation, and is set in the time register BT. Other registers are set to "0". After step S16, a data prefetch routine is called in step S17, and the process returns. If the song play mode is not selected in step S15, the process returns directly. If the switch on which the event occurred is the numeric keypad 21 in step S11, various data setting processes are performed in step S18 depending on the mode, submode, and cursor position at that time. Specifically, setting the style number in pattern mode (display 41 in FIG. 4),
Also, the song number SONG (display 51 in Fig. 5) and song style number STYL (display 51 in Fig. 5) in the song mode.
display 52), tempo TMP, total number of bars NBAR,
Settings for the solo start measure ST, the solo end measure END, and the fill-in measure FLL(i) (where i=0 to FN-1 for the fill-in number FN) are executed according to the input. Then return. If the switch where the event occurred is the right or left cursor key 22c or 22d in step S11, it is determined in step S19 whether the value of the mode register MOD is "1" or "2", that is, the pattern mode or the pattern mode. Determine whether or not it is in edit mode. In any of these modes, step S
20, move the cursor 46 left and right on the screen at that point, for example, to set parameters (excluding styles) in pattern mode (displays 42, 43, and 44 in FIG. 4) or move the cursor in pattern edit mode (displays 42, 43, and 44 in FIG. 4). Display 45) is performed. Then return. If the mode register MOD is not "1" or "2" in step S19, the process returns directly. If the switch in which the event occurred is the rest switch 26 in step S11, it is determined in step S21 whether the value of the mode register MOD is "2" or not, that is, whether the mode is pattern editing mode. In the pattern edit mode, in step S22, rest data is written at the beat position of the pattern on the pattern memory selected based on parameters such as style and quantize and the cursor position (see display 45 in FIG. 4). Return. If the pattern editing mode is not selected in step S21, the process returns directly. Next, referring to the flowchart of FIG. 11,
When the make routine is called, first step S31
Set the song memory pointer PN to "0", fill-in count register FN to "0", and bar number counter BAR to "1".
”, respectively. Next, in step S32, the normal rhythm pattern PTN (STYL, 0, 0) of the intro of style STYL is stored in the normal pattern register BSPT, and the style ST is stored in the fill-in pattern register FLPT.
YL intro fill-in pattern PTN (STYL
, 0, 1) and the solo start bar position ST is set in the intro solo ending boundary pointer FIN. Next, in step S33, an editing routine is called. As a result, the intro part of the song data is automatically generated. Next, in step S34, the normal rhythm pattern PTN (STYL, 1, 0) of the solo section of style STYL is stored in the normal pattern register BSPT, and the fill-in pattern PTN (STYL, 1, 0) of the solo section of style STYL is stored in the fill-in pattern register FLPT. ,1,1),
A value obtained by adding the solo end bar position END and "1" (that is, the ending part start position) is set to the intro solo ending boundary pointer FIN. next,
An editing routine is called in step S35. As a result, the solo section of the song data is automatically generated. Next, in step S36, the normal rhythm pattern PTN (STYL, 2, 0) of the ending part of style STYL is stored in the normal pattern register BSPT, and the normal rhythm pattern PTN (STYL, 2, 0) of the ending part of style STYL is stored in the fill-in pattern register FLPT.
The fill-in pattern of the ending part of PTN (STY
L, 2, 1) are respectively set to the sum of the intro solo ending boundary pointer FIN, the total number of bars of the song NBAR, and "1". Next, step S37
Call the editing routine with . As a result, the ending part of the song data is automatically generated. Finally, in step S38, an end code is set at the final position PART (STYL, PN) of the generated song data, and the process returns. Next, referring to the flowchart of FIG. 12,
When the editing routine is called, first in step S41, the fill-in bar position FLL (FN) specified by the fill-in count register FN is compared with the intro solo ending boundary position FIN. If the fill-in measure position FLL (FN) is smaller than the boundary position FIN,
It is necessary to generate a fill-in pattern within the portion (intro, solo, or ending) that is currently being automatically generated, and the process advances to step S42. In step S42, the value obtained by subtracting the measure position BAR (indicating which measure from the beginning of the song is being processed) from the fill-in measure position FLL (FN) is greater than or equal to "2". Determine whether or not. This value is “2”
”, the normal pattern BSP is used from the measure at this measure position BAR to the measure immediately before the next fill-in pattern.
Since T is to be repeated, the process advances to step S43 to set a repeat pattern. In step S43, fill-in measure position F
Subtract the current bar position BAR from LL (FN) and then
1", that is, the number of times the element of the accompaniment pattern to be repeated at the time of repeat explained in FIG. . Then, in step S44, the element PART(SONG,
Add the repeating start symbol "outside 1" to element PART(S
Normal pattern BSPT to be repeated in ONG, PN+1)
, the repetition end symbol "out 2" and the number of times to go back RTN are added to the element PART (SONG, PN+2), and the element PART (
A fill-in pattern FLPT is set for each of SONG and PN+3). Next, in step S45, "3" is added to the pointer PN pointing to the song memory, and the fill-in number FN is
is incremented to update the current bar position BAR. The normal pattern is repeated RTN+1 times and a fill-in pattern is set, so add the value of register RTN and "2" to bar position BAR, and set the new bar position B.
AR. After step S45, step S41 is performed again.
Return to In step S42, fill-in measure position F
The value obtained by subtracting the current bar position BAR from LL (FN) is "
If the value is not greater than "2", this value is set to "1" in step S46.
Determine whether it is equal to or not. If this value is equal to "1", it means that the bar at this bar position BAR is a normal pattern and the next bar is a fill-in pattern, so the process advances to step S47 to set these patterns. That is, in step S47, the normal pattern BSPT is assigned to element PART (PN) in the song memory, and the fill-in pattern FLPT is assigned to the next element PART (PN+1).
Set each. Next, in step S48, "2" is added to the pointer PN, the fill-in number FN is incremented, and "2" is added to the current number of bars BAR. After step S48, the process returns to step S41 again. In step S46, fill-in measure position F
The value obtained by subtracting the current bar position BAR from LL (FN) is "
If it is not "1" (in this case, this value should be "0"), this means that this bar position BAR is a fill-in pattern, and in step S49 the fill-in pattern is stored in element PART (PN) of the song memory. FLPT
Set. Then, in step S50, the pointer PN,
Fill-in number FN and current minority position BAR,
Increment each. After step S50, the process returns to step S41 again. In step S41, fill-in measure position F
If LL(FN) is greater than or equal to the boundary position FIN, this means that all measures from the current position BAR to this position FIN may be used as a normal pattern, and the process branches to step S51. If the current bar position BAR is equal to the boundary position FIN in step S51, this means that the creation of song data for the relevant portion (intro, solo, or ending) has been completed, and the process returns. If the current bar position BAR is not equal to the boundary position FIN in step S51, the current position BAR is subtracted from the boundary position FIN in step S52, and it is determined whether or not the difference is "1". If the difference is not "1", it is necessary to set the repetition of the normal pattern, so the process advances to step S53. In step S53, a value obtained by subtracting the current position BAR from the boundary position FIN and further subtracting "1", that is, the number of times the pattern returns to the repeating pattern during repeating, is calculated and set in the work register RTN. Then, in step S54, the repetition start symbol "out 1" is added to the element PART (SONG, PN) in the song memory.
+1), the normal pattern BSPT to be repeated in element PA
RT (SONG, PN+2) has a repeat end symbol “outside 2”
and the number of times RTN to go back are set, respectively. Next, in step S55, "2" is added to the pointer PN, the number of times RTN+1 to return to the current bar position BAR is added, and the process returns. [0084] If the difference obtained by subtracting the current position BAR from the boundary position FIN is "1" in step S52, step S5
At step S6, the normal pattern BSPT is set in element PART (PN) of the song memory, and at step S57, the pointer PN and current position BAR are each incremented, and the process returns. Next, referring to the flowchart of FIG. 13,
The data prefetch routine will be explained. The data pre-reading routine reads in advance the accompaniment pattern to be played next when performing automatic accompaniment in song play mode, and sets values in various registers according to the read pattern. In the data prefetch routine, first, in step S61, a counter Q in units of 1/16 beat is cleared to zero. Next, in step S62, the element PART (POINT) of the song memory to be reproduced next is set in the read pattern register PATERN. Then, in step S63, the pointer POINT is incremented. Next, the pattern PATE read out in step S64
Determine whether RN is an end code or not, and if it is an end code, return. If the pattern PATERN is not an end code in step S64, it is determined in step S65 whether the pattern PATERN is the repetition start symbol "outside 1". If the repetition start symbol is "outside 1", the value of the song memory pointer POINT is set in the work register RPST in step S66, and the repeat flub REPE is set.
AT is set to "1" and the process returns to step S62. If the pattern PATERN is not the repetition start symbol "outside 1" in step S65, step S67
Then, it is determined whether the pattern PATERN is the data of the repetition end symbol "outside 2" and the number of times of going back "xn". If it is a repeat end symbol, etc., it is determined in step S68 whether the repeat flag is "1" or not. When the repeat flag REPEAT is "1", it means that the repeat end symbol has been read for the first time after entering the repeat process, so in step S69 the repeat flag REPEAT is
Reset T to ``0'' and store the value obtained by subtracting ``1'' from the previous number of times n to the counter RPN that counts down the number of repeats (at this point, the pattern to be repeated has been read once). , pointer POINT is set to work register RPST (this register RPST contains step S6).
In step S6, the value of (in which the position of the pattern to be repeated is stored) is restored, and the process returns to step S62. [0088] In step S68, the repeat flag REPE
If AT is not "1", this is the second or later repetition end symbol read after the start of the repetition process, so it is determined in step S70 whether or not the down counter RPN is "0". If not "0", the position of the pattern to be repeated in the work register RPST is set in the pointer POINT in step S71, the counter RPN is decremented, and the process returns to step S62. When the counter RPN is "0" in step S70, this means that the last pattern in the repeating process is read ahead, and the process returns to step S62. If the pattern PATERN is not data such as a repetition end symbol in step S67, since it is an accompaniment pattern that has been read out, in step S72
The number of beats of the pattern PATERN is set in the beat register BEET, which sets the beat in units of beats, and the process returns. Next, the interrupt routine will be explained with reference to FIG. The interrupt routine is executed every time there is a timer interrupt (every 1 msec) by the clock 8 (FIG. 1). In the interrupt routine, first in step S81 it is determined whether the counter T is "0". If the counter T is not "0", it means that the time of 1/16 beat has not yet elapsed (that is, the timing of 1/16 beat has not passed), so no sound is generated, and the counter T is decremented in step S82. and return. If the counter T is "0" in step S81, since the timing is every 1/16 beat, the process advances to step S83, and the counter T is set to "0" from the time register BT.
1" is subtracted and the value is stored. The time register BT has 1/
A value representing the time between beats in units of 16 beats in msec is stored in advance, and the value of 1 msec corresponding to the interrupt that led to step S83 is subtracted from the value of this register BT, and the value of the down counter T is calculated. This is the initial value. Next, in step S84, the mode register MOD is set to "5".
In other words, it is determined whether the mode is song play mode or not. If you are not in song play mode, there is no need for automatic performance, so return. When the song play mode is selected, the pattern data PATERN is set in step S85.
(By the data prefetch routine, the register P
ATERN) is the end code and the counter Q is 16 (ie, the end of the measure). If the determination result in step S85 is true, the music has ended, and the process returns directly. If the determination in step S85 is false, the up counter Q is incremented every 1/16 beat in step S86, and the pattern register PA is incremented in step S87.
Rhythm sound ON information and timbre information are output to the tone generator TG in accordance with the information of 1/16×Q-th beat in TERN. As a result, the automatic accompaniment rhythm sound at the timing is generated. In addition, 1 of pattern PATERN
If the information on beat /16×Q is a rest, naturally no sound is produced. Next, in step S88, the beat register BEE
T is multiplied by "4" and it is determined whether or not the product is equal to the value of the counter Q. This is a determination to check whether the measure has come to an end when the automatic accompaniment music is in a time signature other than 4/4 time (such as 3/4 time signature). If the result of this determination is true, the end of the bar has been reached, so in order to pre-read the next pattern data, a data pre-read routine is called in step S89, and the process returns. If the determination in step S88 is false, the process directly returns. As explained above, according to this embodiment, the pattern data is hierarchical as shown in FIG. Data can be automatically generated. [0095] In the above embodiment, the pattern is constructed in units of one measure, but the pattern is not limited to one measure, but may be formed in units of a plurality of measures. Furthermore, the number of bars in each pattern may be different. Furthermore, the pattern data is 1/
The resolution is not limited to 16 beats, but may be configured with various resolutions, and the format is also arbitrary. [0096] In the above embodiment, the pattern is divided by style,
Although patterns are classified in a hierarchical manner using separate indexes for intro solo endings, and separate indexes for normal rhythm and fill-in, the structure for classifying patterns is not limited to this. Also,
The intro and ending may be the same, and there may be several normal rhythms. Furthermore, although the automatic accompaniment device described above is a device that produces rhythm sounds, it is not limited to rhythm sounds; for example, chords or the like may be added to rhythm sounds. [0097] As explained above, according to the automatic accompaniment device of the present invention, accompaniment patterns are stored in a state classified by predetermined indexes, and predetermined parameters including these indexes are stored. Since the song data is automatically generated based on the song data, the song data can be automatically generated based on the general flow of the accompaniment, and the input operation is easy with fewer erroneous inputs.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of an automatic accompaniment device according to an embodiment of the present invention.

[Figure 2] External view of the operation panel

[Figure 3] State transition diagram in this automatic accompaniment device

[Figure 4]
Operation and screen examples of pattern mode and pattern edit mode

[Figure 5] Operation and screen examples of song mode, song make mode, and song play mode

[Figure 6] Conceptual diagram showing the hierarchical structure of pattern data

[Figure 7
] Conceptual diagram showing an example of song structure

[Figure 8] Conceptual diagram showing an example of song data

[Figure 9] Flowchart of main routine

[Figure 10] Flowchart of switch routine

[Figure 11] Flowchart of makeup routine

[Figure 12] Flowchart of editing routine

[Figure 1
3] Flowchart of data prefetch routine

[Figure 14]
Interrupt routine flowchart

[Figure 15] Example of conventional song data

[Explanation of symbols]

1: CPU, 2: Operation panel, 3: Pad, 4: Working memory, 5: Program memory, 6: Pattern memory, 7: Song memory, 8: Clock, 9: Tone generator, 10: Sound system.

Claims (1)

[Claims] 1. A first storage means that stores a plurality of different accompaniment patterns representing the accompaniment content of a certain section in a state classified by a predetermined index, and the accompaniment pattern itself or pattern designation information specifying the accompaniment pattern. a parameter input means for inputting parameters including an index for specifying an accompaniment pattern, and generates the song data according to the parameters input by the parameter input means. An automatic accompaniment device comprising: song data generation means for generating song data.