JPH0876759A - Music device with accompaniment pitch decoding function - Google Patents

Abstract

(57) [Abstract] [Purpose] To provide a music apparatus with an accompaniment pitch decoding function capable of decoding accompaniment sound data expressed in a musical sense to a pitch in accordance with a musical background such as a chord function. [Structure] The accompaniment original data memory 112 stores data in which each sound of the accompaniment is represented by an identifier. When the chord function is given, the CPU 102 converts each chord tone identifier in the original data into a pitch using the chord constituent tone table on the KMAP memory 108. Further, the CPU 102 converts each non-harmonic tone identifier in the original data into a pitch based on the scale database in the SDB memory 108 and the converted harmonic pitch.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a music apparatus having an accompaniment pitch decoding function capable of determining an accompaniment pitch according to a chord function, and can be applied to, for example, an electronic musical instrument having an automatic accompaniment function.

[0002]

2. Description of the Related Art A typical conventional automatic accompaniment device prepares accompaniment original data in which each note of the accompaniment is represented by a pitch (relative pitch) from a chord root note in an accompaniment pattern memory, and the chord is changed by keyboard operation. The root note and type are detected, the pitch data is changed according to the detected chord type, and the detected root note data is added to it to obtain the actual accompaniment pitch. This type of conventional device is a method of determining the accompaniment pitch based on the root note and type of the chord, which is contrary to the musical fact that the accompaniment pitch that can be used cannot be determined unless the tonal harmony function is defined. As a result, this type of device has a problem of producing a musically unnatural sound. In order to solve this problem, Japanese Patent Application Laid-Open No. 3-068922 of the present applicant applies a tonality analysis to a sequence of chords (chord progression) in which each chord is expressed by a root note and a type, and performs the function of each chord (tonality). Harmony function)
We propose a key judgment technique to extract the. In addition, pitch data of each note of the accompaniment is prepared for each chord function as original accompaniment data for automatic accompaniment. This technique reduces the occurrence of unnatural musical sounds.

[0003]

However, there is a problem that a large storage capacity is required in order to prepare pitch data of each accompaniment note for each chord function. Therefore, an object of the present invention is to provide an accompaniment tone whose musical pitch can be determined by musically decoding the pitch of each accompaniment note according to the chord function without preparing pitch data for each chord function in advance. It is to provide a music device with a high decoding function.

[0004]

According to the present invention, the accompaniment raw data storage means for storing the accompaniment raw data storing the accompaniment raw data in which the sequence of the accompaniment tones including the harmony and the non-harmonic tones is expressed by the sequence of the identifiers of the respective sounds, and the chord function A chord function designating means for designating, a chord constituent sound storing means for storing a pitch set of a harmonic sound for the designated chord function, and a scale sound storing means for storing a pitch set of a scale note for the designated chord function; A chord sound decoding means for converting the identifier of each harmony tone in the accompaniment raw data into a pitch using a pitch set of the harmony tone for the specified chord function; and an identifier of each non-harmonic tone in the accompaniment raw data And a non-harmonic sound decoding means for converting the pitch of the harmony sound, which is the conversion result of the harmony sound decoding means, into a pitch based on the pitch set of the scale note for the specified chord function. Characterized Rukoto accompaniment tone pitch decryption function musical apparatus is provided.

According to this structure, each identifier in the accompaniment original data indicates not the pitch of the pitch but the musical function of the pitch. Therefore, in this configuration, accompaniment pitch data for each chord function is not prepared. The identifiers of the respective harmony sounds and the respective nonharmonic sounds in the accompaniment original data are converted into pitches by the harmony sound decoding means and the nonharmonic sound decoding means, respectively, according to the code function. Therefore, a musical accompaniment is possible while saving the storage capacity. By the way, the pitch set of the scale note for the chord function also depends on the musical style.
Therefore, there is a demand for a music apparatus capable of pitch-defining each accompaniment note in consideration of both the music style and the chord function.

To achieve this, according to the present invention, each scale database means stores a plurality of scale database means for storing a pitch set of a scale note for each chord function, and a style selecting means for selecting an accompaniment style. Accompaniment raw data storage means for storing the accompaniment raw data for the selected accompaniment style, which stores the accompaniment raw data in which a string of accompaniment tones including a harmonic sound and a non-harmonic sound is expressed by a string of identifiers of each sound, and Database means for selecting a scale sound database means suitable for the specified accompaniment style from the plurality of scale sound database means, a chord function designating means for designating a chord function, and a chord sound for the designated chord function. A chord component sound storage means for storing a high set, and an identifier of each harmony sound in the accompaniment original data, Harmonic tone decoding means for converting the pitches of the harmony tones for the function, and the identifier of each non-harmonic tone in the accompaniment original data, and the pitch of the harmony tone which is the conversion result of the harmony tone decoding means. And an accompaniment pitch decoding means for converting into a pitch based on the pitch set of the scale sound corresponding to the specified chord function in the selected scale sound database means. A music device with a decoding function is provided.

[0007]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a hardware configuration when the present invention is applied to an electronic musical instrument (automatic accompaniment apparatus) having an automatic accompaniment function. In FIG. 1, a CPU 102 executes a program stored in a program memory 104 to control each part of the device. According to this embodiment, the SDB memory 106 stores multiple scale database for different accompaniment styles. Each scale database contains information about the pitch set of scale notes by chord function.

The KMAP memory 108 stores a plurality of chord component note tables for various accompaniment styles. Each chord component note table contains information on pitch sets of chord notes (chord components) for each chord function.

The accompaniment header memory 110 stores header information for accompaniment patterns of each part for each accompaniment style. As will be described later, in the header information, information indicating which scale sound database in the SDB memory 106 is used, information indicating which chord constituent sound table in the KMAP memory 108 is used, and the accompaniment original data memory 112. Information and the like indicating the accompaniment original data used above are included. The accompaniment original data memory 112 stores accompaniment original data for accompaniment of various accompaniment styles. The accompaniment raw data expresses a string of accompaniment sounds including a harmonic sound and a non-harmonic sound by a string of identifiers of the respective sounds.

The key / function evaluation knowledge memory 114 stores music knowledge data for evaluating the function of each chord in the string of chords (chord progression) in which each chord is represented by the root note and type, and the keynote of the key. Remember. The RAM 116 is used as a work memory for the CPU 102.

The keyboard 118 is a performance data input device. In the automatic accompaniment mode, the CPs for the chords sequentially designated by the key depression pattern in the left hand area of the keyboard 118
U102 detects the root note and type of the designated chord. In this way, a string of chords (chord progression) in which each chord is represented by a root note and a type is obtained. For this chord progression,
The CPU 102 refers to the music knowledge data stored in the key / function evaluation knowledge memory 114 and executes key / function determination processing to determine the keynote and chord function of each chord. As the adjustment / function determination process, Japanese Patent Application No.
The one disclosed in No. -68922 can be used.

The accompaniment style selector 120 is an input device for selecting an accompaniment style. Sound source 122 is CP
It is a device that electronically generates a musical tone signal for a melody and accompaniment under the control of U102. Sound system 12
Reference numeral 4 reproduces the tone signal from the sound source 122 as a sound.

The automatic accompaniment apparatus has an accompaniment original data memory 11
2 has a pitch decoding function of musically decoding each identifier of the accompaniment sound based on a given accompaniment style and chord function and converting it into pitch data (representing a pitch for a keynote). The accompaniment style input from the selector 120 is used to specify the accompaniment raw data, the scale database (SDB), and the chord component sound table to be used via the accompaniment header information.

When the chord function is designated, the pitch decoding function first uses the pitch set of the harmony tone corresponding to the designated chord function in which the identifier of each harmony tone in the accompaniment original data is written in the chord constituent tone table. Convert to data. After that, the pitch decoding function uses each non-harmonic sound identifier in the accompaniment original data to decode the pitch data of the harmonic sound and the pitch data of the scale sound for the specified chord function written in the scale database. Convert to.

Further, the automatic accompaniment apparatus has not only the above-mentioned pitch decoding function but also a correction function for musically testing the pitch decoding result and correcting it in a timely manner. The first correction function corrects the shape formed by the pitch train according to the musical situation.
For this purpose, a shape adjustment section is prepared, and in the case of a predetermined chord function, the pitch string in the shape adjustment section is corrected according to the change data included in the accompaniment original data. The second correction function examines whether or not the pitch decoding result of the non-harmonic sound is a musically unfavorable avoid sound, and corrects the accompaniment sound to a pitch other than the avoid sound if applicable.

FIGS. 2 and 3 show an example of the scale database by reference numeral 106E. In these figures, the leftmost column shows the code function (eg I). As shown in the figure, the scale sound database 106E has information on the pitch set of the scale sound for each chord function. For example, in the case of the chord function I in the first line, the scale notes are C, D, E, F, G,
It is composed of A and B. In the figure, C is a symbol ST1 indicating the first scale, and D is a symbol S indicating the second scale.
T2, and so on, ST3 for E, ST4 for F,
The symbols ST5 for G, ST6 for A, and ST7 for B are attached. Furthermore, among these scale notes, C, E, G
Is also a chord constituent sound (harmonic sound), so K
The symbols 1, K2 and K3 are attached. Further, since F is an avoid sound, the symbol AV is added. In other words, the scale sound database of the embodiment also serves as a table of avoid sounds. Furthermore, the scale database 106E has an arrow.

[Outer 1] , →, ← symbols are entered. this house

[External 1] indicates that the scale note has the same strength to combine with the upper (right) and lower (left) adjacent scale notes. The → indicates that it is easy to combine with the upper scale, and the ← indicates that it is easy to combine with the lower scale.

Note that note names C to B shown in FIGS. 2 and 3 actually represent note names when the keynote is C, in other words, the pitch from the keynote (relative pitch to the keynote), and C From the keynote once (0 in numerical expression), C
^# (D ^b ) is minor second (1) from the keynote, and so on. Further, the numbers given to ST and K are for convenience and are not always necessary. For example, in the case of the chord function I, it is sufficient to show that C (the same note name as the keynote) is a scale note and a harmony note. Therefore, a scale database with a data format of 5 bits per 1 note name, that is, 1 bit indicating whether it is a scale note, 1 bit indicating whether it is a harmony note, 1 bit indicating whether it is an avoid note, and 2 bits indicating a combination Can be configured.

FIG. 4 shows an example of the chord component sound table with reference numeral 108E. The chord constituent table has information on pitch sets of chord sounds (chord constituent notes) for each chord function. The chord component sound table 108E shown in the drawing collectively shows tables for 3 poly (K1 to K3) and 4 poly (K1 to K4). For example, in the case of the chord function I, K3 is C4 (C in the fourth octave), K2 is E4, K3 is G4 in 3 poly, and K4 in 4 poly.
C5 is added as. The chord component sound table 10
The pitch data of C4 or the like in 8E represents the pitch of the harmony sound when the keynote is C, in other words, the pitch from the keynote.

The chord component sound table 108 shown in the figure is also included.
E is a table for the "adjacent" mode that maintains the accompaniment tone range as much as possible regardless of the chord function. For this reason, the arrangement of the harmony tones for some chord functions has an inversion form in which the lowest note K1 is not the root note. For example code function IV
C3, which is originally the third component tone, is assigned to K1 as a three polyphonic tone for, and F4 (chord root tone) is provided as K2, and A4 (originally the second component tone) is provided as K3. In contrast to the chord component note table for the "adjacent" mode, a chord component note table for the "parallel" mode is also prepared, which shifts the accompaniment tone range in parallel according to the root note of the chord.

FIG. 5 shows an example of a header for the accompaniment part with reference numeral 110E. The header memory 110E stores the address pointer of the scale database used for the symbol address "SDB". In addition, the symbol address "KMA
The address pointer of the chord component note table to be used is stored in P ″. By these, the scale note database and the chord component note table used for pitch decoding of each note identifier of the accompaniment original data are specified. , “Style”, “Beat”, and “Pattern length” are respectively stored with accompaniment style, beat, and pattern length information.

The symbol address "number of lines" stores the number of lines of the accompaniment that composes the accompaniment part, and the addresses of the subsequent lines store the address pointers of each line data in the accompaniment original data memory 112. .
In the symbol address "AVT", the permissible length of a void sound (for example, the length of a sixteenth note) is stored. As described below,
When an avoid sound exceeding the allowable length is generated, the avoid sound is removed. Section information for adjusting the shape of the pitch string of the accompaniment is stored in the symbol address "shape adjustment section" (NUL is set for an accompaniment having no such section).
L value is set).

FIG. 6 shows an example of the accompaniment original data memory forming the accompaniment part with reference numeral 112E. In the illustrated example, the accompaniment part consists of three lines. Each line data starts from the symbol address "line 1", "line 2", "line 3", respectively. At the first address of the line 1, the code K3 indicating the third code tone is stored as the identifier of the harmony tone. At the next address, the address pointer of the rhythm is stored. In the rhythm memory pointed to by the rhythm address pointer, as shown at 212, a sounding timing pattern and a mute timing pattern of the accompaniment sound are written. At the next address, the end code END of the line is stored. Therefore, line 1 consists only of the accompaniment sound of K3. Similarly, the line 2 consists only of the accompaniment sound of K2 (second harmonic sound).

On the other hand, the line 3 has a sound (A) and K1.
It is composed of an accompaniment sound of (first harmony sound). Therefore, line 3
At the first address of the above, a code A that means a sound is stored as an identifier of the non-harmonic sound that is the first accompaniment sound. At the next address, the basic tonal harmony mode, that is, the pitch data of the low note A when the keynote is "C" and the chord function is "I" is stored. In this embodiment, the pitch data in the basic tonal harmony mode is stored together with the identifier of the non-harmonic sound on the accompaniment original data memory. At the next address, the storage address pointer of the rhythm data is stored as the rhythm information of the sound A.
Further, a code K1 indicating the first harmony tone is stored at the next address as an identifier of the harmony tone, and the harmony tone K1 is stored at the next address.
The address pointer of the rhythm information is stored.

The accompaniment part indicated by the accompaniment original data in FIG. 6 has a pattern length of one bar, and each line is formed in the format shown in FIG. In this way, the accompaniment original data represents each sound of each accompaniment sound including a harmony sound and a non-harmonic sound by a string of identifiers. Therefore, for automatic accompaniment, it is necessary to convert each identifier into pitch data for a given chord function using the scale note database and chord constituent note table designated by the header.

In FIG. 8, (A) is a musical score together with the rhythm of the pitch decoding result when the key is “C” and the chord function is I (the chord name is C _Major ) with _respect to the accompaniment raw data of FIG. It is shown. In the case of code function I, K3 of line 1 = G4, K2 of line 2 = E4, A of line 3
= D4, K1 = C4. FIG. 8 (B)
_Shows the pitch decoding result when the key is "C" and the chord function is IV (the chord name is F _Major ) for the same accompaniment original data. In case of code function IV, K3 of line 1 =
A4, K2 of line 2 = F4, A of line 3 = D4, K
The pitch is converted to 1 = C4.

FIG. 9 is a flowchart of the chord function-based accompaniment pattern creating process 9-G executed by the automatic accompaniment apparatus. The purpose of this process 9-G is to decode each note identifier in the accompaniment original data for each chord function to determine its pitch. In this embodiment, the accompaniment pattern creation process 9-G for each chord function is performed as shown in step 9-0.
This is executed in response to an accompaniment style selection operation by the accompaniment style selector 120. In the first step 9-1 of the accompaniment pattern creation process for each chord function, the CPU 102 reads the header associated with the selected accompaniment style. Then, by using the scale database address pointer written in the header, the scale database SDB matching the selected accompaniment style is accessed and the first chord function of the SDB is selected (9-2).

Thereafter, in 9-3 to 9-12, accompaniment pitch decoding processing for the selected chord function is executed. First, in step 9-3, the CPU 102 selects the first accompaniment line by setting the pointer for scanning the accompaniment original data to the value of the first line address pointer written in the header. Then, in 9-4 to 9-10, the pitch decoding process for the selected accompaniment line is executed. First, in the chord pitch decoding step 9-4, the CPU 102 converts the pitch of each chord note included in the selected accompaniment line into a pitch using the chord constituent note table. Next, it is checked whether or not the selected accompaniment line includes a shape adjustment section and the selected chord function is a chord having four chords, that is, a chord having four constituent notes (9-5). Correct the shape of the accompaniment line.

Next, in step 9-7, it is checked whether or not the selected accompaniment line contains a non-harmonic sound, and if it is applicable, a non-harmonic sound decoding process 9-8 is executed to identify each non-harmonic sound in the line. To pitch. Next, in a void check / correction step 9-9, the result of the non-harmonic sound decoding process is tested, and if an undesired avoid sound, that is, an avoid sound exceeding the allowable length AVT is generated, the sound is not a void sound. Change to high. In step 9-10, it is checked whether the pitch decoding processing has been completed for all the lines, and if not completed, the next accompaniment line is selected (9-1
Return to 1) and 9-4.

At step 9-12, it is checked whether the pitch decoding processing is completed for all chord functions. If not, the next chord function is selected from SDB (9-13), and 9-3 is entered. Return. In this way, in the accompaniment pattern processing 9-G for each chord function, the accompaniment pitch sequence of each line of the accompaniment part is decided for each chord function, and the result is stored in the RAM 11.
Stored in a predetermined area of 6.

After that, when the performance starts on the keyboard 118, C
The PU 102 detects the type and root note of the chord from the key depression pattern in the left-hand key range in the chord determination process, and discriminates the function of the chord and the keynote in the tonality analysis process. Then, the accompaniment process is executed, the accompaniment pitch of each line of the accompaniment part for the determined chord function is read from the RAM 116, and the keynote data is added to the corresponding accompaniment pitch data from the RAM 116 at each sounding timing to generate the actual sound. Produces high data, thereby controlling the sound source 122. This allows
Automatic accompaniment that matches the accompaniment style and chord function is realized.

FIG. 10 shows a flowchart of the harmony sound decoding process 9-4. First, in 10-1, the CPU 102 advances the scan pointer of the accompaniment original data from the line start address to locate the first chord tone identifier of the accompaniment line.
Next, at 10-2, the CPU 102 looks up the chord constituent tone table KMAP pointed to by the header and extracts the pitch data for the corresponding chord tone identifier in the selected chord function. For example, if the harmony tone identifier of the accompaniment original data is K3, the chord function is II _m , and the KMAP used is the 4-poly adjacent chord component note table 108E shown in FIG. 4, the 4-poly K3 column of this table and II _m The pitch data A4 indicated at the intersection with the line is extracted as the pitch decoding result of K3.

After the pitch of one harmonic tone identifier is decoded, the accompaniment original data scanning pointer is advanced to locate the next harmonic tone identifier (10-4), and the pitch decoding 1
Perform 0-2. When the accompaniment original data scan pointer reaches the line end code END (10-3), the processing is completed. The shape correction processing 9-6 realizes a function of adjusting the shape of the pitch string of the accompaniment original data. This shape correction processing 9-6 is executed for the shape adjustment section when the accompaniment pattern has a shape adjustment section and the chord function is a four chord. For this reason, the header has information regarding the presence or absence of a shape adjustment section and the area, and the accompaniment original data memory also has information for shape adjustment as necessary.

In FIG. 11, an example of accompaniment original data whose shape is adjusted is partially shown by a reference numeral 112G. Change data is placed next to the chord identifier as shown. The change data "0" indicates that the harmony pitch data decoded in the harmony pitch decoding processing 9-4 is not changed, and the change data "+1" changes the decoded harmony pitch data to the next higher harmonic sound. The change data “−1” means that the decoded chord pitch data is changed to the next lower chord sound.

FIG. 12 is a flowchart of the shape correction processing 9-6. First, in 12-1, the CPU 102 locates the first chord sound identifier from the original data memory of the accompaniment line. It is checked whether or not the harmony tone identifier located in 12-2 is a sound within the shape adjustment section. This is performed by comparing the shape adjustment section information placed in the header with the rhythm information of the chord sound identifier.

If the sound is in the shape adjustment section, the flow proceeds to 12-3, and the pitch data which is the pitch decoding result for the located harmony sound identifier is read from the RAM 116. Next 12
At -4, the change data for the located harmony tone identifier is read from the accompaniment original data memory 112. When this change data has a value other than "0" (12-5), 12-
Proceed to step 6 to convert the change data into a pitch value. This is obtained by calculating the pitch difference (pitch) between the chord notes for the selected chord function from the chord constituent note table. For example, assuming that the 4-poly adjacent table 108E shown in FIG. 4 is used as the chord constituent tone table KMAP, the chord function is V7, the chord tone identifier is K4, and the change data is “+1”, the decoding result for K4 is B4 (numerical value). 4x in expression
12 + 11 = 59). Regarding code function V7,
The pitch between K4 and the next higher harmonic is the difference between the pitch name B of K4 and the pitch name D of K1 (12 mod (2-11 + 1
2)), that is, the minor third (3 in numerical expression). Therefore, adding a minor third to B4 results in D5 (62 = 59 + 3), which determines the changed harmonic pitch (12-7).

Then, until the end of the accompaniment line (12-8), the identifier of the harmony tone is found from the original data (12-
By performing the processing shown in 12-2 to 12-7 for each 9), the shape of the harmonic pitch sequence in the shape adjustment section is corrected. For example, the identifier of the harmony sound in the shape adjustment section is K3-K.
4-K4-K3, the changed data string is 0-0-1-1,
Assuming that the chord composition note table KMAP used is the 4-poly / adjacent type in FIG. 4, when the chord function is I, the pitch string G4-
C5-C5-G4 is obtained. When the chord function is I _M7 , the pitch sequence G4-B4-B is obtained as a result of the chord decoding process 9-4.
4-G4 is formed. Further, by the shape correction processing 9-6, this pitch string is changed to G4-B4-C5-B4.

FIG. 13 is a flowchart of the non-harmonic pitch decoding process 9-8. First, the original data scan pointer is advanced from the line start position to locate the first non-harmonic tone identifier (13-1). The type of the non-harmonic sound identifier located in 13-2 is determined. In the present embodiment, the low sound A, the transitional sound P and the lost sound E are used as the non-harmonic sound identifiers.
Therefore, the sound deciphering process 13-3, the elapsed sound deciphering process 13-4, and the lost sound deciphering process 13 are performed in accordance with the determined nonharmonic sound type.
-5 is executed to determine the pitch of the non-harmonic sound.

After the pitch decoding process is performed for one non-harmonic tone identifier, the original data scanning pointer is advanced to search for the next non-harmonic tone identifier, and if found (13-
7), and similar pitch decoding processing 13-2 to 13-5 is repeated. When the original data scan pointer reaches the line end code END, the processing is completed.

FIG. 14 shows a flowchart of the sound decoding process 13-3. First, in 14-1, the CPU 102 checks whether or not the chord function is the standard chord function I, and if so, the process proceeds to 14-2, and the original pitch data is read (14-3). Next, the pitch data of the chord following the chorus (which has already been obtained by the chord deciphering process 9-4 and the shape correction process 9-6) is read (14-4), and the chord is changed to the succeeding chord. The direction DIR of the pitch of is calculated (14-4). The spur direction information DIR is used for pitch decoding of the spur for other chord functions.

As described above, in the embodiment, the traveling direction of the sound from the low-pitched sound to the subsequent chord sound is the same for all chord functions. That is, for chord functions other than the reference chord function (14-1), the pitch data of the following chord sound is read (14-5) and the scale database S
The pitch of a scale note (which may be a chord at the same time) adjacent to the pitch of the subsequent chord and having a direction of DIR is read as a sound from DB (14-6).

For example, if the chord function = N, the pitch of the subsequent chord sound = C4 = 12 × 4, SDB = the scale database 106E of FIG. 2, DIR = “-” (downward), then CP
U102 is a chord function IV of the scale database 106E
From the column of the pitch class C of the chord to the right. Then, it is found that the pitch class D is a scale note. Therefore, the pitch of the low-pitched sound is longer than the second harmonic tone by 2 degrees or more (indicated by a value 2 moved to the right from C), and the low-pitched pitch = D4 (= 12 × 4 + 2).

FIG. 15 shows a flowchart of the elapsed sound decoding processing 13-4. When the code function is the standard function I (1
5-1) Copy the elapsed pitch data written in the original data memory (15-2). 1 for other code functions
Proceeding to 5-3, the harmonic pitch data before and after the elapsed sound is read (15-3), and the pitch is compared (15-4). When the pitches of the preceding and following chord sounds are different, the procedure proceeds to 15-5, and the pitch data of the elapsed sound is initialized to the pitch of the preceding harmony sound. Then (15
-6), the scale note database SDB is searched from the pitch of the previous harmonic sound to the pitch of the subsequent harmonic sound, and every time a scale sound is found, the elapsed pitch data is updated to that pitch.

Decoding examples are shown in Tables 1 and 2.

[Table 1]

[Table 2]

On the other hand, if the preceding and following chord tones have the same pitch, the flow proceeds to 15-7, and the coupling degree data of the upper and lower scale tones adjacent to the pitch of the chord tones are extracted from the scale database SBD. Then, if there is a scale note having a stronger degree of coupling, it is set as the pitch of the elapsed note. When both coupling degree data are the same or the coupling degree data is not written, the lower scale tone is set as the pitch of the elapsed tone.

Table 3 shows a decryption example

[Table 3]

FIG. 16 shows a flowchart of the lost sound decoding processing 13-5. When the code function is the standard function I (16
-1), copy the pitch loss data from the original data memory (16-2). Further, the pitch data of the subsequent harmony tone is read (16-3), and the pitch INT of the missing tone for the subsequent harmony tone is obtained (16-4). This pitch information INT is used for all chord functions to make the jump progression from the missing sound to the subsequent chord sound similar. That is, for the chord functions other than the standard function, the following pitch data of the harmony tone is read (16-4), the tone scale closest to (INT + chord tone pitch) is found from the scale database, and this is searched. The pitch of the lost sound is set (16-5).

As described above, in the non-harmonic sound decoding processing, the identifier of each non-harmonic sound is converted into a pitch based on the pitch decoding result of the harmonic sound and the scale database. In the above example, the pitch data of the non-harmonic sound corresponding to the reference chord function is provided in the accompaniment original data memory, but a method without it is also possible. This can be realized by finely dividing the meanings of the non-harmonic sounds (types of the non-harmonic sounds identifiers) and performing a decoding process according to each meaning.

For example, the sound A is divided into a sound "+ A" which is a downward sound and a sound "-A" which is an ascending sound with respect to the subsequent harmony sound, and is + A.
In decoding (-A), SDB is searched in a direction higher (lower direction) than the subsequent harmony sound. Further, regarding the progressed sounds, those having different pitches of the preceding and following harmony tones are P groups, and those having the same pitch of the preceding and following harmony tones are the N groups. Further, regarding the progressed sound of the P group, when two scales are included between the preceding and following harmonys on the SDB, the sound defined by the scale closer to the preceding harmony is P1, and the scale closer to the latter harmony. The sound defined by is differentiated by P2. Further, the N group is divided into + N, which has a meaning as an upper scale tone of the front and rear chords, and −N, which has a meaning as a lower scale tone of the front and rear chords.

The lost sound is expressed by an identifier including pitch information for the following harmony sound. For example, E (+5)
Is a missing note identifier having a meaning as a scale note that is five semitones higher or is closest to the following harmonic sound. In this way, by enriching the non-harmonic tone identifiers and performing processing according to their respective meanings, it is possible to perform more detailed pitch decoding and also to generate non-harmonic pitch data from the accompaniment original data for the reference chord function. Can be removed.

FIG. 17 shows the void check correction processing 9-9.
The flowchart of is shown. First, in 17-1, the original data operation pointer is advanced from the accompanying line start position to locate the first non-harmonic tone identifier. In 17-2, the non-harmonic sound decoding processing result, that is, the pitch data, for the located non-harmonic sound identifier is read. Then, the pitch class and chord function are opened, the scale sound database SDB is looked up, and it is tested whether the decoded pitch is an avoid sound (17-3). If it is an avoid sound, it is checked whether the sound length exceeds the allowable length AVT written in the header (FIG. 5). If this is the case, it is a musically unfavorable avoid sound, so the avoid removal processing 17-4 is executed to change the sound to a sound that is not an avoid sound.

After that, each time a non-harmonic tone identifier is found from the original data of the accompaniment line (17-7), the above process is repeated, and when the original data pointer reaches the line end code END (17-6), the process is exited.

FIG. 18 is a flowchart of the void removal processing 17-5. First, in 18-1, the pitches of the harmony sound before and after the avoid sound are extracted. Next, (18-2), the type of non-harmonic sound decoded as an avoid sound is identified. When the non-harmonic sound is a dip A, the scale database SDB is referenced to find a scale note that is adjacent to the pitch of the subsequent harmony note and is on the opposite side of the avoid sound, and corrects it. (18-3). In the case of the missing note E, the scale note database SDB is referred to, a scale note adjacent to the avoid note is found, and it is set as a corrected pitch (18-4). In the case of the elapsed sound P, the harmonic pitches before and after are compared, and if they are the same (18-
5) Then, the sound is regarded as the low sound A, and the low sound correction processing of 18-3 is executed. If they are different, the scale note database SDB is referred to, and the scale notes other than the avoid notes between the preceding and following chord notes are searched. If such a scale note is found, it is set as a corrected pitch of the elapsed note. If not found (18-
7), it is regarded as a dull sound, and a dull sound correction process 18-3 is executed.

By performing the void check / correction processing in this manner, it is possible to realize an accompaniment that does not generate a musically unfavorable avoid sound. It is preferable in terms of processing speed that the avoid check / correction process is performed each time one non-harmonic sound is decoded by the non-harmonic sound decoding process. Although the description of the embodiment has been completed, various modifications can be made within the scope of the present invention.

[0054]

As described in detail above, according to the present invention, the accompaniment raw data is provided with an identifier representing the musical meaning of each note, and the chord component note storage means and the scale note are assigned according to the designated chord function. Since the pitch of each note is decoded by referring to the database means, it is not necessary to store the pitch data for each chord function in advance, and a musically natural accompaniment can be realized. Further, according to the present invention, the scale note database means adapted to the accompaniment style is used at the time of pitch decoding, so that the accompaniment in the desired style can be secured.

[Brief description of drawings]

FIG. 1 is a block diagram showing a hardware configuration of a music apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a part of a scale sound database.

FIG. 3 is a diagram showing a part of a scale sound database.

FIG. 4 is a view showing an example of a chord component sound table.

FIG. 5 is a diagram showing an example of an accompaniment header memory.

FIG. 6 is a diagram showing an example of an accompaniment original data memory.

FIG. 7 is a diagram showing a sequence of sound identifiers of three accompaniment lines.

FIG. 8 is a diagram showing an example of an accompaniment in musical score.

FIG. 9 is a flowchart of an accompaniment pattern creating process for performing pitch decoding for each chord function.

FIG. 10 is a flowchart of a harmonic pitch decoding process.

FIG. 11 is a diagram showing accompaniment original data in which the shape of a pitch example is modified.

FIG. 12 is a flowchart of shape correction processing.

FIG. 13 is a flowchart of non-harmonic pitch decoding processing.

FIG. 14 is a flow chart of the sound decoding process.

FIG. 15 is a flowchart of a process of decoding elapsed sounds.

FIG. 16 is a flow chart of the sound decoding process.

FIG. 17 is a flowchart of a void check / correction process.

FIG. 18 is a flowchart of a void removal process.

[Explanation of symbols]

 102 CPU 104 Program memory 106 SDB (scale note database) memory 108 KMAP (chord constituent tone table) memory 110 Accompaniment header memory 112 Accompaniment original data memory 120 Accompaniment style selector

Claims (2)

[Claims] 1. An accompaniment raw data storage unit for storing original accompaniment data in which a string of accompaniment tones including a harmony tone and a non-harmonic tone is expressed by a sequence of identifiers of each tone, and a chord function designation unit for designating a chord function. A chord component sound storage means for storing a pitch set of a harmony tone for a designated chord function, a scale tone storage means for storing a pitch set of a scale tone for the designated chord function, and in the accompaniment raw data. Harmonic sound decoding means for converting the identifier of each harmony sound into a pitch using the pitch set of the harmonic sound for the designated chord function; and the identifier of each non-harmonic sound in the accompaniment original data as the harmony sound decoding means. And a non-harmonic sound decoding means for converting the pitch of the harmonic sound that is the conversion result of the above into a pitch based on the pitch set of the scale note corresponding to the specified chord function, That accompaniment pitch decoding function with music equipment. 2. A plurality of scale database means in which each scale database means stores a pitch set of a scale note for each chord function, a style selecting means for selecting an accompaniment style, and accompaniment original data for the selected accompaniment style. The accompaniment source data storage means for storing accompaniment source data in which a sequence of accompaniment tones including a harmony tone and a non-harmonic tone is expressed by a sequence of identifiers of each tone, and a scale tone suitable for the selected accompaniment style. Database selection means for selecting the database means from the plurality of scale note database means, chord function designation means for designating chord functions, and chord constituent note memory for storing pitch sets of chord sounds for the designated chord functions Means, and a pitch set of the harmony tone corresponding to the designated chord function, the identifier of each harmony tone in the accompaniment original data Harmonic tone decoding means for converting to pitch using, the identifier of each non-harmonic sound in the accompaniment original data, the pitch of the harmony sound which is the conversion result of the harmony sound decoding means, and the selected scale database A non-harmonic sound decoding means for converting a pitch based on a pitch set of a scale note corresponding to the designated chord function in the means, and a music apparatus with an accompaniment pitch decoding function.