JP3554772B2 - Obrigado Automatic Generator - Google Patents

Description

[0001]
[Industrial applications]
The present invention relates to a music apparatus, and more particularly to an automatic obligade generation apparatus that automatically generates obligado (a counter melody compared to a melody).
[0002]
[Prior art]
In the conventional obligade automatic generation apparatus, one chord component sound is selected for each chord as an obligade sound, thereby generating an obligade line. According to this conventional technique, the amount of data to be stored is small, but there is a problem that the obligade as a generation result becomes monotonous.
On the other hand, in an automatic accompaniment apparatus for attaching an accompaniment to a melody, accompaniment pattern data is prepared and stored in advance, and the pitch of each sound of the accompaniment pattern data is determined according to the detected chord. Usually, several types of patterns such as normal, variation, intro, and fill-in are prepared as accompaniment patterns, and any one of the patterns is selected according to the progress of music. For example, in the automatic accompaniment device disclosed in Japanese Patent Application Laid-Open No. 62-25792, the pattern is switched from a normal pattern to a variation pattern when the key depression time of a melody key exceeds a predetermined time.
In any case, this type of automatic accompaniment apparatus has a problem that a considerable storage capacity is required for accompaniment pattern data. In addition, since the accompaniment pattern does not have the character of obligade, it is difficult to apply the accompaniment pattern to the automatic obligade generation device.
[0003]
[Problems to be solved by the invention]
Accordingly, an object of the present invention is to provide an automatic obligor generation apparatus capable of generating an obligade that can be changed in various ways in accordance with the progress of a melody without requiring a large amount of data.
[0004]
[Means, action]
According to the present invention, the chord progression storage means for storing the chord progression, the melody storage means for storing the melody, and, for each chord section in the chord progression, one chord component sound is assigned as a kernel sound to obliquely store the chord. multiple nuclear generating means for generating a nucleus and expansion section detecting means for detecting a section where melody tone does not change during Jo Tokoro as a developing section obbligato from the melody, the Kakuon to the detected expanded section Expansion unit generating means for generating an obligade expansion unit expanded to a melody composed of a sound of sound, and means for generating a sound train of the obligade by combining the generated core of the obligade and the obligade expansion unit. An automatic obligade generation device, comprising:
[0005]
According to this configuration, while the melody is moving as a melody, the oblique sound is generated according to the rhythm of chord progression by the nuclear sound generated by the nucleus generating means. On the other hand, in a section in which the melody is not moving, an obligato expansion section is generated by expanding the nuclear sound into a melody composed of a plurality of sounds by the operation of the expansion section detection means and the expansion section generation means. In this way, an obligato that moves in contrast to the melody can be obtained. Furthermore, the automatic obligade generation device does not need to previously store pitch sequence data such as accompaniment pattern data, so that storage capacity can be saved.
In one configuration example, the development unit generation means includes: a development unit generation rule storage unit expressed by a plurality of tone types; a scale sound storage unit that stores scale sound information for each function of a chord; And a pitch decoding means for converting the kernel tone type of the expansion unit generation rule into a pitch in accordance with the stored scale note information for the chord function in.
According to this configuration, it is possible to include scale sounds other than the chord constituent sounds in the obligato expansion unit. In addition, the generated obrigado developing section has a pitch content according to the function of the ongoing chord (tonal harmony function), so that musical naturalness that does not deviate from the key can be secured.
[0006]
【Example】
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 shows a hardware configuration of the automatic obligor generation apparatus according to the embodiment. The CPU 10 executes a program stored in the program memory 12 to control each unit of the device. The work memory 14 is used as a work memory of the CPU 10 in a RAM configuration. The input device 16 includes a keyboard, from which a melody, a chord progression, a start sound of an obligato, and the like are input. The melody memory 18 stores the melody data input from the input device 16. The chord progression memory 20 stores chord progression data input from the input device 16. Here, the chord progression data has information on the function (for example, V7) and length of each chord in the chord progression, and the note name and length of each key. The chord function and the key information are chord names represented by chord types and chords obtained by chord detection from chord keypress data input from the input device 16 (for example, chord names represented by type 7 and chord G). G7) can be obtained by key and function analysis. A technique disclosed in Japanese Patent Application No. 3-68922 can be used as this kind of key / function analysis.
[0007]
According to the present embodiment, the SDB memory 22 stores scale note information for each chord function. The CPU 10 refers to this scale sound information when selecting a chord constituent sound as the core sound of the obligade, and when deciphering each tone type in the development part of the obligade.
According to the present embodiment, the expansion rule memory 24 stores a sequence of a plurality of sound types as rules for generating an expansion of an obligato.
[0008]
Obligado consists of a combination of Obrigard's core and deployment. The CPU 10 stores the generated obligated data in the obligated memory 26.
At the time of reproduction, the CPU 10 reads out sound information from the melody memory 18 and the obligator memory 24, controls the sound source 28, and generates a melody sound from the sound source 28 and generates a tone signal of an obligator sound. The generated tone signal is emitted through a sound system 30 including an amplifier and a speaker.
[0009]
FIG. 2 shows a functional block diagram of the automatic obligor generation device of the embodiment. In the first stage, the automatic obligade generation apparatus generates a core of the obligade (which forms the obligade where the melody is moving) in the first stage, and develops the obligade development unit (the melody is moving) in the second stage. That form the obligado where it does not exist).
[0010]
For this purpose, the nucleus generation unit 102 receives the obligade start sound given from the obligade start sound input unit 160 as an obligato sound (nuclear sound) for the first chord section, and after the second chord, one sound per chord. The nucleus of obligado is generated by determining the nuclear sound at a rate. Here, chord constituent sounds are used as nuclear sounds. For this purpose, the nucleus generator 102 reads the function and key information of each code in the chord progression from the chord progression memory 20. Then, by looking up the SDB memory 22 using the chord function, the pitch data of the chord component sound in the key C is selected from the SDB memory 22, and the key data is added thereto to determine the pitch of the nuclear sound.
[0011]
On the other hand, the development section detection unit 104 examines the melody in the melody memory 18 and does not move the melody, that is, the melody sound is over a predetermined period (for example, four beats) from a predetermined time (for example, the beginning of a bar). A section that has not changed is detected as an expanded section of the obligade. The generation rule selection unit 106 selects an expansion unit generation rule to be applied to the detected expansion section from the rule base of the expansion unit rule memory 24. The rule decoding unit 108 converts each tone type written in the selected development unit generation rule into a pitch to generate expansion unit data. For this pitch conversion, the rule interpretation unit 108 reads the key and the chord function in the development section from the chord progression memory 20, reads the scale note information for the chord function from the SDB memory 22, and among them, the scale indicated by the note type. Find the pitch and add a key to it to determine the pitch of the note type.
The synthesizing unit 110 synthesizes obligated nuclear data from the nucleus generating unit 102 and expanded data from the rule decoding unit 108 to obtain final obligated data.
[0012]
FIG. 3 shows an example of the chromatic note database SDB stored in the SDB memory 22. As shown in the figure, the chromatic note database stores chromatic note pitch information (pitch information when the key is C) for each chord function (example I, IIm). The symbol "K" indicates that the pitch is a scale tone of a chord constituent sound, and the symbol "ST" indicates that the pitch is a scale note other than the chord constituent sound. For example, in the case of the chord function I, it is shown that C, E, and G are chord constituent sounds K and D, A, and B are chromatic notes ST other than chord constituent sounds.
[0013]
FIG. 4 shows an example of a rule base stored in the expansion unit rule memory 24. The first rule base DEV1 stores a set of expansion unit rules applicable to the expansion section from the beginning to the end of the bar. In the figure, t1, t2, t3, and t4 represent the first beat, the second beat, the third beat, and the fourth beat, respectively. For example, the first rule of the rule base DEV1 is that the first beat is a tone type SD (i), the second beat is a tone type K +, the third beat is a tone type K +, and the fourth beat is a tone type SD (i). Means Here, SD (i) indicates an obligato nuclear sound in the development section, and K + indicates a K sound adjacent above the previous sound. The meaning of each sound type will be described later. The second rule base DEV2 stores a set of expansion section rules applicable to the expansion section from the second beat to the end of the bar, and the third rule base DEV3 stores two or three beats from the beginning of the bar. A set of expansion section rules applicable to the expansion section from the beat to the end of the bar is stored.
[0014]
As shown in FIG. 5, the CPU 10 generates an oblique guard nucleus in the first stage (5-1), and generates a development unit in the second stage (5-2).
[0015]
FIG. 6 shows a flowchart of the nucleation process 5-1. First (6-1), the CPU 10 puts an end mark at the end of the chord progression. Next, the obligade start pitch data input from the input device 16 is read as the obligade nuclear sound for the first chord section (6-2). Initialize to i = 0 (6-3), set the obligator start pitch data to SD (i), and set the first code length data to SL (i). Here, the arrays SD (0), SD (1),... Represent the pitch sequence of the obligado in the obligator memory 26, and the arrays SL (0), SL (1). Express. Note that the obligato start sound is selected from among the constituent sounds of the first chord.
[0016]
Subsequently, the content of SD (i) is set in the previous obligatory pitch register OB (6-5), i is incremented, and the function and key of the next chord are read from the chord progression memory 20 (6-6). If the end mark has not been reached, a constituent tone K of the current chord closest to the previous obligatory pitch register OB is found from the scale note database. For example, assuming that the previous obligatory pitch = B4 = 4 × 12 + 11, the current chord function = V7, and the current key = G = 7, “C” in FIG. 3 is read as “G” and “E” in FIG. The column is set as a search start position corresponding to B4. If you move one to the right, you will find the chord composition sound of V7. Therefore, the pitch of the nuclear sound for the current chord section is determined as K = B4 + 1 = C5 = 5 × 12. The determined nuclear pitch K is set to SD (i), the current chord length is set to SL (i) (6-9), and the process returns to 6-5.
By continuing the above processing until the end mark is reached, an obligated sound (nuclear sound) is generated at a rate of one sound per chord.
[0017]
FIG. 7 shows a flowchart of the development unit generation processing 5-2.
The CPU 10 checks the melody from the beginning, and searches for a place where the melody sound is extended by two or more beats under the same chord (7-1). If found, it is checked whether or not the melody includes a portion extending four beats from the beginning of the bar (7-2). In this case, since the section of four beats from the beginning of this bar is the development section of the obligade, the process proceeds to 7-3, and one development part generation rule is selected from the rule base DEV1 of four beats. Then, each tone type of the rule is converted into a pitch to generate expanded portion data (7-4), which is inserted into the obligade data arrays SD (i) and SL (i) to synthesize obligade. (7-5).
[0018]
If it does not correspond to 7-2, it is checked at 7-6 whether the melody sound includes a portion extending from the second beat to the end of the bar. If this is the case, the section from the second beat to the end of the measure is an expanded section of an obligade having a length of 3 beats, so go to 7-7 and proceed to one expanded section from the rule base DEV2 of 3 beats. Select a generation rule. Then, each tone type of the rule is converted into a pitch to generate expanded section data (7-8), and inserted into the obligor data arrays SD (i) and SL (i) to synthesize obligade (7-). 9).
[0019]
If it does not correspond to 7-6, it is checked in 7-10 whether the melody sound includes a portion extending from the third beat to the end of the measure or extending from the first beat to two beats. If so, since it is an obligato expansion section having a length of two beats, the process proceeds to 7-11, and one expansion part generation rule is selected from the rule base DEV3 of two beats. Then, each tone type of the rule is converted to a pitch to generate expanded section data (7-12), and inserted into the obligor data arrays SD (i) and SL (i) to synthesize obligade.
By repeating the above processing until the end of the melody (7-14), a developed part is generated and combined with the core of the obligade to complete the obligade.
[0020]
FIG. 8 shows the details of the pitch decoding process for decoding the tone type of the expansion rule. The tone type SD (i) is converted into a pitch by reading the pitch data of the nuclear sound for the current chord from the obligato kernel data.
[0021]
The tone type K + (or K-) is converted into a pitch by reading an adjacent K tone above (or below) the pitch of the preceding obligato tone from the scale note database SDB. The tone type ST + (ST−) is converted into a pitch by reading an ST sound adjacent to (above) the preceding obligato sound from the scale sound database SDB. ST + (SD (i + 1)) (or ST− (SD (i + 1))) is converted into a pitch by reading an ST sound adjacent to (above) the nuclear sound in the next chord section from the chromatic note database SDB. ST + _(t i) (or ST- _(t i)) is converted to the pitch by reading the ST sound adjacent the upper (lower) than i beat of obbligato sound from the scale sound database SDB.
[0022]
The 4-beat synthesizing process 7-5 is performed as follows. Now, the expanded part pitch sequence is D (1), D (2), D (3), D (4), and the pitch sequence is L (1), L (2), L (3), L (4). (The length is one beat each), the length from the beginning of the current kernel sound SD (i), SL (i) to the beginning of the development section is L (s), and the end of the development section to the current kernel sound is L (s). Assuming that the length to the end is L (e), SD (i) = SD (i), SL (i) = L (s), SD (i + p) = D (p) (where p = 1, 2 , 3, 4), SL (i + p) = L (p), SD (i + 5) = SD (i), and SL (i + 5) = L (e). However, when L (s) = 0 or L (p) = 0, the corresponding data is unnecessary. Other combining processes 7-9 and 7-13 can be executed in the same manner.
[0023]
FIG. 9 shows an operation example of the automatic obligor generation device of the embodiment. Code VI m in the first beat of the first bar as shown as chord progression in (b), V, 2 becomes VI m in bars after progress is given in the second beat to fourth beats. The key is C through one or two measures as shown in FIG. On the other hand, the automatic obligor generation device executes the nucleus generation process 5-1 to generate the oblique nucleus shown in (d). Here nucleus E5 in the section first code VI m, D5 in section of the next code V, the E5 in the section last code VI m. Next, the automatic obligor generation device executes a developing unit generation process 5-2. Here obbligato automatic generation system detects that it extends the same melody tones A4 under the same code VI m in the second measure th as shown in 200. Here, A4 is the first to third beats and the fourth beat is rest, but during the rest, the melody is assumed to have no change, and the section of four beats from the beginning to the end of the second bar is Obligade It is detected as a development section. Here, it is assumed that the automatic obligator generation apparatus selects rule 2 from the expansion part generation rule base DEV1 of four beats (t1 to t4). According to the rule 2, the first beat t1 is a tone type SD (i) indicating the present nuclear sound, the second beat t2 is a chord constituent sound K adjacent to the preceding tone (in this case, SD (i)), and the third beat. t3 is a scale note that is not a chord constituent sound adjacent to SD (i) and the fourth beat t4 is SD (i). Therefore, when this rule 2 is decoded using the chord function VIm, the key C, and the chromatic note database SDB, an obligato expansion section as shown in (e) is obtained. Here, the obligato development section forms a melody in which the first beat is the pitch E5 of the current nuclear sound, the second beat is A5, the third beat is F5, and the fourth beat is E5. Finally, by combining the nuclear line (d) and the developed part (e), an obligato as shown in (f) is completed.
[0024]
The description of the embodiments has been finished above, but various modifications are possible within the scope of the present invention. For example, in the embodiment, the expansion unit generation rule base is for quadruple, but the expansion unit rule base of another time signature (for example, 3 time) is prepared, and the expansion unit rule base corresponding to the time signature of the music to be used is set to the expansion unit rule. You may make it select from the memory 24.
[0025]
【The invention's effect】
As described in detail above, according to the present invention, the nucleus generation means allocates one chord constituent sound for each chord section as an obligade sound (nuclear sound) to generate a nucleus of obligade, and the development section detection means A section (development section) in which the obligado is to be developed is detected from the stationary area of the melody, and an obligade expansion section is generated by expanding the nuclear sound into a melody composed of a plurality of sounds with respect to this expansion section by the expansion section generation means. Therefore, an obligator line as opposed to the movement of the melody line can be generated without requiring a large amount of data.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a hardware configuration of an automatic obligor generation device according to an embodiment;
FIG. 2 is a block diagram showing an obligade generation function of the automatic obligade generation device according to the embodiment;
FIG. 3 is a diagram showing a scale sound database.
FIG. 4 is a diagram illustrating a generation rule base of an obligato expansion unit.
FIG. 5 is a flowchart of an obligato generation process.
FIG. 6 is a flowchart of a nucleation process.
FIG. 7 is a flowchart of a development unit generation process.
FIG. 8 is an explanatory diagram of a developed pitch decoding process.
FIG. 9 is an explanatory diagram of the operation of the embodiment.
[Explanation of symbols]
18 melody 20 chord progression memory 102 nucleation unit 104 development section detection unit 106 generation rule selection unit 108 rule interpretation unit

Claims (2)

Chord progression storage means for storing chord progression;
Melody storage means for storing a melody;
Nucleus generation means for allocating one chord component sound as a kernel sound for each chord section in the chord progression to generate an obligato nucleus;
And the expansion section detecting means for detecting a section in which melody sound does not change the period of Jo Tokoro as a developing section of obbligato from the melody,
Expansion unit generating means for generating an obligato expansion unit which expands the nuclear sound into a melody composed of a plurality of sounds with respect to the detected expansion section,
Means for combining the generated nucleus of obligade and the obligade expansion unit to generate a sound string of obligado,
An automatic obligade generation device, comprising: The automatic obligato generation apparatus according to claim 1,
The expansion unit generation means includes:
Expansion unit generation rule storage means for storing an expansion unit generation rule represented by a plurality of strings of sound types;
Scale sound storage means for storing scale sound information for each function of a chord,
Pitch decoding means for converting each tone of the expansion unit generation rule into a pitch in accordance with the stored scale note information for the function of the chord in the detected expansion section,
An automatic obligade generation device, comprising:

Images