JPH0527750A - Automatic accompaniment method - Google Patents

Abstract

PURPOSE:To automatically generate accompaniment data conforming to a played melody. CONSTITUTION:Interval and sound volume data on a sound played on a musical instrument 101 are inputted to a central processing unit 103 through a sound input device 102. A cell state update routine 104 determines the transition rule of the state change of a cell and updates the state of the cell. An accompaniment data output routine 105 determines the interval of accompaniment data according to the state of the cell obtained by the cell state update routine 104 and a sound input device 102 sends the accompaniment data to an acoustic signal generation device 107 by using the inputted sound volume data to make an automatic accompaniment conforming to the performance of a player. Further, a melody difficulty detection routine 106 uses the frequency of update of the transition rule to detect melody difficulty which is an index showing how easy the understanding of the played melody is.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic accompaniment method capable of automatically generating accompaniment data and outputting an accompaniment in synchronization with a performer, and analyzing the melody of a played music to detect the melody difficulty level. Melody difficulty detection method.

[0002]

2. Description of the Related Art As discussed in Japanese Patent Laid-Open No. 57-8596, a conventional automatic accompaniment method is a method in which an accompaniment pattern input in advance is read out at any time and is output in accordance with a performance.

Further, as discussed in Japanese Patent Laid-Open No. 59-58485, in the conventional melody difficulty level detecting method, whether the key corresponding to the note length and the note played is a white key or a black key. Is a method of detecting the melody difficulty level by using.

[0004]

In the above-mentioned conventional automatic accompaniment method, since the accompaniment pattern input in advance is read and used, there is a problem that only a monotone accompaniment output can be obtained and a variety of automatic accompaniment systems cannot be realized. It was

Further, in the conventional melody difficulty level detecting method, only the difficulty level of performance is detected and it is not an index showing the difficulty level of understanding when listening to music.

An object of the present invention is to provide an automatic accompaniment method capable of automatically generating various accompaniment data.

Another object of the present invention is to provide a melody difficulty level detecting method capable of obtaining not only a difficulty level at the time of playing but also an index indicating a difficulty level at the time of listening to music.

[0008]

[Means for Solving the Problems] A cell corresponding to each keyboard is set and each state is represented by a binary value of 1 or 0. Each time a sound input signal is sent to the central processing unit, the cell state transition rule is determined so that the cell corresponding to the pitch becomes 1 and the cell separated by a certain distance from this cell becomes 0. Furthermore, the state of the cell is updated according to this transition rule, and accompaniment data is automatically generated using the pitch indicated by the cell that continuously takes a value of 1 from the history of the states of all cells. Further, the above object is achieved by repeating the processing of determining the melody difficulty level by using the number when the transition rule is different from the transition rule determined at the time of the previous sound input.

[0009]

[Function] By imitating the process of information processing in the human brain in a simple manner, the transition rule of the state of cells corresponding to nerve cells in the brain is determined for each sound input, and this transition rule is used for all Update the cell state of. In addition, accompaniment data is generated using the history of cell states updated according to this transition rule. As a result, due to the nature of the cellular automaton, it is possible to output an accompaniment that reflects the dynamic characteristics of the played melody ascending / descending, in synchronization with the performer, realizing a variety of automatic accompaniment systems. it can. Also,
By using a different frequency for the transition rule determined at the time of sound input and the transition rule determined at the time of the previous sound input, the difficulty of interpreting the melody of music can be imitated at the nerve cell level in the brain. It is possible to determine the difficulty level of the tuned melody.

[0010]

[Example] When sound is emitted, the vibration of the sound wave causes the vibration of the basement membrane inside the cochlear duct of the inner ear as a traveling wave through the eardrum and the ossicle. The distance that the traveling wave reaches the basilar membrane differs depending on the frequency of the sound, with high sounds remaining near the entrance to the cochlear canal and low sounds reaching the upper part. Hair cells are in contact with each position of the basement membrane and are converted into an electric signal and transmitted to the auditory cortex of the cerebrum. It is said that each nerve cell of the cerebrum constantly changes its state while affecting each other. In addition, the transition rules for state changes of nerve cells are constantly updated. By using numerical values for the state of nerve cells and simple algebra for the transition rules of state changes, it is possible to mimic the state of the brain when listening to music. This makes it possible to automatically generate comfortable accompaniment data and obtain the degree of difficulty in understanding music.

An embodiment of the present invention will be described below with reference to the drawings. In FIG. 1, 101, 102, 107, 108 are devices, 104,
105 and 106 are programs built into the computer. When the sound played on the instrument (101) is taken in from a microphone, the result of frequency analysis via the sound input device (102) is used, and in the case of a keyboard instrument, the key pressing position and key pressing speed are displayed. , Pitch and volume data, and fetched in the central processing unit (103). Cell status update routine (10
In 4), the transition rule of the state change of the cell is determined, and all the states of the cell are updated according to this transition rule. In the accompaniment data output routine (105), the cell state update routine (10
The pitch of the accompaniment data determined based on the state of the cell obtained in 4) and the volume of the accompaniment data obtained by using the volume data captured by the sound input device (102) are compared to the acoustic signal generator (107 ) To perform automatic accompaniment according to the performance of the performer. Also, the melody difficulty detection routine (10
In 6), the transition rule update frequency is obtained using the number of cases where the transition rule determined in the cell state update routine (104) is different from the one determined at the time of the previous sound input, and the transition rule update frequency is obtained. It is used as the melody difficulty, which is an index showing the ease of understanding. Further, this is output to a melody difficulty output device (108) such as a computer screen or a printer.

Next, a cell state update routine (104), an accompaniment data output routine (105), which form the central part of this embodiment,
The processing procedure of the melody difficulty detection routine (106) will be described. In the following, for simplicity, it is assumed that the performance is performed by a single note.

First, a cell used in the following process, its state, and a transition rule will be described. N cells corresponding to each keyboard are set, and each state is represented by a binary value of 1 or 0. It is assumed that the state of each cell changes from the state of itself and the cells on both sides at the time of the previous sound input according to a certain transition rule. Ct (i) ∈ {0, 1}
Let (i = 0, 1, ..., N-1) be the state of the i-th cell at the time of the sound input at the t-th time. Also, let Rt be the transition rule of the cell state. The state Ct + 1 (i) of the i-th cell at the time of sound input for the (t + 1) th time is Ct (i-
1), Ct (i), Ct (i + 1) are used to determine as follows.

[0014]

[Equation 1]

Since the argument that Rt takes is three variables represented by 0 or 1, determining Rt is equivalent to determining the value (0 or 1) for a set of 8 types of arguments. These are defined as Ak ∈ {0, 1} (k = 0, 1, ..., 7) as follows.

[0016]

[Equation 2]

In the following, the cell state is stored in the cell state table, and the transition rule is stored in the transition rule table in the form of Ak (k = 0, 1, ..., 7). In addition, the transition rule table is set as these initial settings at the start of performance.
Ak = 0 (k = 0, 1, ..., 7) and the initial state of the cell
C1 (i) (i = 0, 1, ..., N-1) is played in the first measure Pj (j = 1, 2, ..., Jmax; Jmax is played in the first measure The number of sounds that are played) is set as follows. For all i (i = 0, 1, ..., N-1),

[0018]

[Equation 3]

The process 104 in FIG. 1 will be described below. As shown in FIG. 2, each time a sound is input, a transition rule is determined (steps 201 to 204) and the cell state is updated (steps 205 to 210).

In step 201, the position of the cell corresponding to the frequency of the sound played at the t-th time is set to g (g = 0, 1,
..., N-1). As shown in FIG. 5, the transition rule is updated so that the g-th cell is 1 (292-2) and the cells on both sides are 0 (292-1, 292-3).
Perform steps 2 to 204.

The state of the g-2, g-1, g-th cell is referred to from the cell state table 1 (step 292), and Ct (g-
2), Ct (g-1), and Ct (g) are stored (step 202-1). Define x as

[0022]

[Equation 4]

Transition rule table 1 so that Ax = 0
Rewrite (291) (step 202-2). Similarly, g, g + 1,
The state of the g + 2nd cell is referred from the cell state table 1 (292) and stored in Ct (g), Ct (g + 1), Ct (g + 2), respectively.
(Step 203-1). Define z as

[0024]

[Equation 5]

Transition rule table 1 so that Az = 0
Rewrite (291) (step 203-2). Furthermore, g-1, g,
The state of the g + 1th cell is referenced from the cell state table 1 (292) and stored in Ct (g-1), Ct (g), Ct (g + 1), respectively.
(Step 204-1). Define y as

[0026]

[Equation 6]

Transition rule table 1 so that Ay = 1
Rewrite (291) (step 204-2). As described above, the transition rule can be updated at the time of the t-th sound input.

Next, the cell state updating method (steps 205 to 210) will be described. First, the cell state table 1 storing the cell states updated at the time of the previous sound input.
All the contents of (292) are set in the cell state table 2 (29
3) is copied (step 205). After substituting 1 for counter i (step 206), until i becomes N-1 or more
(Step 210), the following steps 207, 208 and 209 are repeated.

In step 207, the state of the i-1, i, i + 1th cell is referred to from the cell state table 2 (293),
Store them in Ct (i-1), Ct (i), and Ct (i + 1), respectively. Define w as

[0030]

[Equation 7]

The cell state table 2 (293) is rewritten by referring to Aw from the transition rule table 1 (291) as the state Ct + 1 (i) of the i-th cell (step 208). The i is counted up (step 209), and if i is smaller than N-1, the process returns to step 207 and the operation is repeated. When i becomes N-1 or more by these operations, the states of all cells are written in the cell state table 1 (292).

Next, a method of generating the pitch of the accompaniment data in the processing 105 of FIG. 1 will be described with reference to FIG. After substituting 0 for counter i (step 301), i is N
Until the above (step 306), the following steps 302, 3
Repeat steps 03, 304, and 305. First, the state of the i-th cell updated at the time of the previous and current sound input is stored in two cell state tables (292 (cell state at the time of this sound input),
Read from 293 (cell state at the time of the last sound input),
They are stored in Ct + 1 (i) and Ct (i), respectively (step 302).
When Ct (i) = 1 and Ct + 1 (i) = 1 in step 303, the operation of the next step 304-1 is performed, and otherwise, the operation of step 304-2 is performed. In step 304-1, the i-th cell firing duration S (i) is counted up and the cell firing duration table (391) is rewritten. In step 304-2, 0 is substituted for the i-th cell firing duration S (i) to rewrite the cell firing duration table (391). Count up i (step 305),
If i is smaller than N, the process returns to step 302 and the above operation is repeated (step 306).

In step 307, referring to the cell firing duration table (391), Gp cells having a large value among S (i) are selected and used as pitches of accompaniment data. Here, Gp is a value indicating how many accompaniment sounds are output for one sound input. In step 308, of the sound input data read from the sound input device (FIG. 1: 102), the volume data is stored in Vin. Using this, set the volume of accompaniment data as follows:
(Step 309).

[0034]

[Equation 8]

However, w (0 ≦ w, real number) is a preset coefficient. The above expression means that the accompaniment volume is output in proportion to the volume of the input, so that, for example, a quiet accompaniment can be output for a quiet performance, and an accompaniment that matches the nature of the performance is obtained. be able to.

As described above, by outputting the pitch / volume accompaniment data obtained in steps 307 and 309 to the acoustic signal generator (FIG. 1: 107) (step 310), automatic accompaniment is realized.

Next, a method of determining the difficulty level of the melody in the process 106 of FIG. 1 will be described with reference to FIG. First, in step 401, the contents of the transition rule table 1 (291) updated by the cell state update routine (FIG. 1: 104 and FIG. 2) and the contents of the transition rule table 2 (491) updated at the last sound input. To match. As a result, when the transition rule at the previous sound input does not match (step 402), the following steps 403 and 404 are performed.

In step 403, the transition rule update number counter (Ms, 492) is counted up, and in step 404, all the contents of the transition rule table 1 (291) are copied to the transition rule table 2 (491). Then step 405
In, the sound input number counter (Mt, 493) is incremented. Further, in step 406, the transition rule update frequency Q is calculated by the following equation with reference to the transition rule update frequency counter (Ms, 492) and the sound input frequency counter (Mt, 493),

[0039]

[Equation 9]

Q is output as the melody difficulty level to the melody difficulty level output device (FIG. 1: 108) (step 407). This mimics the frequency of controlling the state transition rules of nerve cells in the brain.

By repeating the above processing, accompaniment data can be generated at any time, and various accompaniment data can be automatically generated. Further, the difficulty level Q (Equation 9) of the played melody can be determined at any time for the played melody, and an index indicating the difficulty level of the music can be obtained.

According to the present invention, it is possible to determine the difficulty level of understanding a melody using the played melody and automatically generate accompaniment data, so that various functions including automatic accompaniment with the performer are realized. can do. Further, in music education, the melody difficulty level can be used as an index when practicing instrumental or vocal music by determining the difficulty level of understanding the melody using the melody of music.

Further, according to the present invention, since the above functions can be realized by a simple circuit with a small processing amount, it can be incorporated in a device using a microcomputer.
It is possible to realize an automatic accompaniment system that operates in synchronization with the performer in real time.

[0044]

According to the present invention, it is possible to determine the difficulty level of understanding a melody using the played melody and automatically generate accompaniment data. Function can be realized.

[Brief description of drawings]

FIG. 1 is a block diagram of an embodiment of an automatic accompaniment method of the present invention.

FIG. 2 is a flowchart of a cell state update routine according to the embodiment of this invention.

FIG. 3 is a flowchart of an accompaniment data output routine according to the embodiment of this invention.

FIG. 4 is a flow chart of a melody difficulty detection routine according to the embodiment of the present invention.

FIG. 5 is a conceptual diagram of a transition rule updating method according to the embodiment of this invention.

Claims (5)

[Claims] 1. A cell corresponding to each keyboard is set,
Each time the sound input signal is taken into the central processing unit, the position data pressed is used to determine the transition rule of the cell state,
An automatic accompaniment method characterized in that the states of all cells are updated in accordance with this transition rule, and accompaniment scale data is generated using the history of cell states. 2. The state of the cell corresponding to each keyboard is set to 0.
It is represented by a binary value of 1 or 1, and the transition rule of the cell state is determined so that the state of the cell corresponding to the depressed position is 1 and the state of the cell at a certain distance from this cell is 0. ,
An automatic accompaniment method characterized in that the states of all cells are updated in accordance with this transition rule, and accompaniment scale data is generated using the history of cell states. 3. Each time a sound input signal is taken into the central processing unit, a transition rule of the state of a cell represented by a binary value of 0 or 1 corresponding to each keyboard is determined, and according to this transition rule. An automatic accompaniment method, characterized in that a pitch corresponding to a cell that continuously takes a value of 1 is used as accompaniment data in the updated history of cell states. 4. Playing accompaniment data obtained by using a history of cell states updated for each sound input corresponding to each keyboard by combining with a system for converting accompaniment data into an acoustic signal. 2. The automatic accompaniment method according to claim 1, wherein the automatic accompaniment method is performed in synchronization with the output. 5. A cell corresponding to each keyboard is set,
Each time a sound input signal is captured, the key position data is used to determine the transition rule of the cell state, and it is determined whether this transition rule is equal to the transition rule determined at the time of the previous sound input. A method for detecting melody difficulty, which comprises obtaining the update frequency of a transition rule and then determining the difficulty level of understanding the melody.