JPH05346781A - Key detecting device and automatic music arranging device - Google Patents

Abstract

PURPOSE:To enable automatic music arrangement without any inconvenience even in the case of modulation by detecting a modulated key and a modulation position even if music data contains modulation data and arranging music according to the key and position. CONSTITUTION:A disk controller 2 to which a floppy disk device 1 is connected, a CPU 3, a program memory 4, a working memory 5, a parameter memory 6, a melody memory 7, a key memory 8, a priority order table 9, an accompaniment data memory 10, a chord memory 11, a pattern memory 12, etc., are connected to one another through a bus line 14. Then a key detecting means analyzes music data stored in the music data storage means 7 and detects the key and modulation position. The key storage means 8 is stored with the key and its variation state so that they correspond to the position of the music data. The music data generating means 3 generates musical performance information on a performance part other than data stored in the music data storage means 7 according to the stored key data and the modulation position in the case of the modulation, and performs automatic music arrangement.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a key detecting device for detecting a key from music information such as a melody, and an automatic music arranger for performing a music arrangement on the music information.

[0002]

2. Description of the Related Art Conventionally, for the purpose of automatically arranging music information such as a melody, a technique for investigating the key of the music information, which is the premise, has been studied. For example, in all keys, the ratio of the pitch name of the scale note of the key and the pitch name of the constituent notes of the melody is calculated, and the key with the highest degree of matching is used as the key. Are known.

[0003]

However, when the underlying music information includes a key modulation, it is not possible to know the fact that the key has been modulated, and the wrong key is detected in return. There was a problem that inconvenient automatic arrangement was performed.

An object of the present invention according to claim 1 is to enable automatic arrangement without inconvenience even when there is a modulation.

Another object of the present invention is to enable accurate detection of the modulation position.

[0006]

In order to achieve the above-mentioned object, the invention according to claim 1 is a music data storage means for storing music data, and a key of the music data and its key are analyzed by analyzing the music data. Key detecting means for detecting a key position when a key is being transposed, and key storing means for storing key data representing the key, which corresponds to a key that changes sequentially when the music data is keyed. Storing the key data corresponding to the position of the music data, and storing the key corresponding to the key data stored in the key storage means, the key changing sequentially when the key is transposed. Music data generating means for generating other music data to be played together with the music data based on the position of the music data being played.

According to a second aspect of the present invention, music data supplying means for supplying music data including note data, section specifying means for sequentially specifying a predetermined section of the music data, and notes within the specified predetermined section. A coincidence degree calculating means for obtaining a coincidence degree between the note name of the data and the note name of the in-tone scale, and an identifying means for sequentially identifying the predetermined section so as to have an overlapping portion with an adjacent section are sequentially identified. And a modulation unit that detects a modulation in accordance with the degree of coincidence that is sequentially calculated for the section.

[0008]

According to the first aspect of the invention, the key detecting means analyzes the music data stored in the music data storing means to detect the key and the transposition position. The key storage means stores the key and its change state in correspondence with the position of the music data. The music data generation means generates performance information of performance parts other than those stored in the music data storage means based on the stored key data and the modulation position when there is a modulation, and performs automatic arrangement.

According to the second aspect of the present invention, the degree of coincidence between the note data and the intone is calculated for each predetermined section, and the key is selected according to the degree of coincidence. The predetermined section, which is a unit for calculating the degree of coincidence, is sequentially specified so as to have an overlapping portion with the adjacent section, and the degree of coincidence is sequentially calculated accordingly. Modulation is detected based on the change in the degree of coincidence.

[0010]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the overall structure of an automatic arrangement device including the structure of a key detection device according to an embodiment of the present invention. In the figure, a disk controller 2 to which a floppy disk device 1 is connected, a central processing unit (CPU)
3, program memory 4, working memory 5, parameter memory 6, melody memory 7, key memory 8, priority table 9, accompaniment data memory 10, chord memory 1
1, the pattern memory 12 and the output interface 13 are connected to each other via a bus line 14.

The floppy disk device 1 reads the musical score data of the melody recorded on the floppy disk and supplies it to the melody memory 7 via the disk control device 2. As shown in FIG. 2A, the musical score data is a combination of a key code MKCm indicating a pitch and a duration time MDTm (m = 0, 1, 2, ...) Indicates a duration (hereinafter referred to as “note data”). , M is “note data number”
,) And a bar line code indicating the break of a bar, and an end code indicating the end of the melody, and are stored in the melody memory 7 in the illustrated format.

The melody memory 7 is a RAM (Random).
Access Memory) and is used to increase the processing speed.

The CPU 3 controls the reading of the note data, detects the key based on the read note data, and arranges the note data. The program memory 4 and the parameter memory 6 are ROM (Read O
nly Memory), which stores a program for operating the CPU 3 and necessary control parameters. The working memory 5 is a RAM, and C
PU3 is used for temporary storage of data during calculation.

The key memory 8 is a RAM and stores keys determined by the CPU 3. As shown in FIG. 2B, the data stored in the key memory 8 (hereinafter referred to as “key data”) is bar number MEj, key note TNj, and mode (major or minor) MDj (j = 1, 2). , 3, ..., Hereinafter referred to as “key data number”). However, the data of j = 0, that is, the first data is composed of only the tonic TN0 and the mode MD0 because the measure number is unnecessary.

The priority table 9 is a table in which the keys (proximity keys) that are likely to be transposed as shown in FIG. 3 are arranged in the order of high possibility, and are stored in the ROM. In the same figure, for example, if the first tone is C major (C), then F
Most likely to change to major (F), then G
Major (G), A minor (Am), ..., C minor (Cm).

This table is used to accurately detect modulation in key detection, as will be described later.

Accompaniment data memory 10 and chord memory 11
Is a RAM, which is used for temporary storage of chords and bar numbers assigned to each bar during execution of automatic arrangement, and temporary storage of note data of the selected and modified accompaniment pattern. ..

The pattern memory 12 is a ROM, in which data of accompaniment patterns corresponding to pattern numbers is stored for each part such as bass, and is used as basic data at the time of automatic arrangement.

FIG. 4 is a diagram showing the note name data structure of the in-tone (diatonic note) of each tone, with the tone number n = 0.
C major (C) and A minor (Am), n = 1 corresponds to C # major (C #) and B ♭ minor (B ♭ m) respectively, and n = 11 B major (B) and G #. Up to the minor (B # m), 24 tones are assigned corresponding to n = 0 to 11. SCALE the note name data
When represented by n (k), for example, SCALE0
(0) = 0 and SCALE1 (3) = 5. The value of this note name data is the major note of C major, that is, SCALE0
The value of (0) is set to 0, and the value of 1 is added each time a semitone is raised to set the value of other note name data. However, the value 12 is the value 0
Return, that is, the remainder value divided by the value 12 is used (hereinafter, when such a value is used, it is indicated by adding "mod12").

In the present embodiment, the note name data SC of this intone is
The degree of coincidence between ALen (k) and the note name of the melody to be detected is calculated for each detection section, and key detection is performed based on this degree of coincidence. For example, as shown in FIG. 9, four bars are set as one detection section, and the detection sections are moved one bar at a time to calculate the degree of coincidence, that is, the adjacent detection sections are moved so as to have an overlapping portion of three bars. The degree of coincidence is calculated. In the example of FIG. 9, the section length ME of the detection section
N is 4 and shift length (movement amount of the detection section) SME is 1.

The contents of the key detection processing executed by the CPU 3 will be described below with reference to FIGS.

FIG. 5 is a flowchart of the main routine of the key detection process and the automatic arrangement process.

In step SM1, the key detection subroutine shown in FIG. 6 is executed, and in step SM2, the key detection subroutine shown in FIG.
Execute the automatic arrangement subroutine shown in step SM3
Then, the contents of the key memory 8 in which the key detection result is written are recorded on the floppy disk, and this routine is ended.

FIG. 6 shows the step SM1 of the main routine.
7 is a flowchart of a subroutine of the key detection processing shown by.

In step S1, the musical score data recorded on the floppy disk is written in the melody memory 7,
Then, the bar number i at the beginning of the detection section is set to the value 1 (step S2), and the initial tone detection subroutine shown in FIG. 7 is executed (step S3).

In step S4, the bar number i at the beginning of the detected section is incremented by the shift length SME, and the key data number j (corresponding to j in FIG. 2B) is set to the value 1.

Next, the coincidence degree calculation subroutine shown in FIG. 8 and the key judgment subroutine shown in FIG. 8 are executed (steps S5 and S6), the measure number i is incremented by the shift length SME (step S7), and (i + MEN-). It is determined whether or not 1) exceeds the total number of measures TME (step S).
8). If (i + MEN-1) does not exceed TME (the answer in step S8 is negative (NO)), the process returns to step S5, and if it exceeds (the answer in step S8 is affirmative (YE).
S)), and this routine ends.

FIG. 7 is a flow chart of the initial tone detection subroutine executed in step S3 of FIG. 6. In step S10, the coincidence degree calculation subroutine is executed as in step S5 of FIG. Therefore, first, the contents of the matching degree calculation subroutine shown in FIG. 8 will be described.

In step S31 of FIG. 8, the length of the detected section length MEN represented by the number of measures is converted into a length with an eighth note as the value 1, and the converted value is set as the converted section length SUM. In the example of FIG. 10, the section length of 4 bars corresponds to 32 eighth notes, so SUM = 32.

Next, the key number n is set to 0 (step S3).
2) The first note data number m0 of the first measure (measure number i) of the inspection section is set to the initial value of the note data number m, and the duration time of the note data that matches the in-tone is integrated. The matching data length SAME which is a parameter for initialization is initialized to a value of 0 (step S33). For example, in the example of FIG. 11, when i = 1, m0
Since m = 0, m = 0, and when i = 3, m0 = 11
Therefore, m = 11.

In step S34, the key code MKCm
Is expressed by mod12, that is, MKCmmod12 obtained by converting pitch data into pitch name data and the pitch name data SCALEn (0) to SCALEn (6) (Fig. 4) of the internal tone are discriminated. If there is no match, the process immediately proceeds to step S36. If there is a match, the duration time MDTm is added to the match data length SAME (step S35), and the process proceeds to step S36. In step S36, the note data number m is incremented by the value 1, and whether or not the key code MKCm is data within the detection section currently targeted for key detection (measure number i to (i
It is determined whether the data is within the section up to + MEN-1)) (step S37). When this answer is affirmative (YES), the process returns to step S34, and when negative (NO), the degree of coincidence RATE2 (n) is calculated by the following equation (1).

RATE2 (n) = SAME / SUM (1) Next, the key number n is incremented by the value 1 (step S39), and if n = 12 is not satisfied (the answer to step S40 is negative (NO)), Returning to step S33, if n = 12 (the answer to step S40 is affirmative (YES)), this routine ends.

According to the subroutine of FIG. 8, the degree of coincidence L between the note name of the note data and the note name of the intone within one detection section
RATE2 (n) (n = 0 to 11) is calculated. Since the coincidence RATE2 is calculated by dividing the coincidence data length SAME with an eighth note as a unit by the section length SUM (Equation (1)), the coincidence degree weighted by the note length is given. Becomes

Returning to FIG. 7, in step S10, the degree of coincidence L
After calculating RATE2 (n), the RATE2 (n) is calculated.
Let the value of (n) be RATE1 (n) (step S1
2). RATE1 (n) is a parameter provided for observing the tendency of change in the degree of coincidence in the key determination subroutine of FIG. 9 described later. In this routine, LRATE1
(N) and LRATE2 (n) have the same value, so L
Processing is performed using RATE1 (n).

In step S13, LRATE1 (0)
~ N of LRATE1 (n) which is the maximum among (11)
Is set as the selection candidate key number LSCLN (p), and the number of n is set as the selection candidate key number LN. This takes into consideration that there may be a plurality of keys with the highest degree of coincidence, and the parameter p takes an integer value from 0 to (LN-1). That is, for example, if the key numbers of the key with the maximum degree of coincidence LRATE1 are 0 and 5, LN = 2 and P are 0 and 1.
Therefore, LSCLN (0) = 0, LSCLN (1) =
It becomes 5.

In step S13, the selection candidate key number LS
CLN (p) is divided into a major key number LTN (p) and a minor key number LTN (p + LN). This is because, as shown in FIG. 4, one major key and one minor key correspond to one key number n (for example, n = 0 corresponds to C major and A minor). Here, the major key number LTN (p) is the selection candidate key number LSCLN (p).
Is used as is, minor key number LTN (p + LN)
Is set to {LSCLN (p) +9} mod12.
This is because the minor tone having the same internal tone as that of the major tone takes into account the tone which is nine semitones higher.

LMD (p) is a mode parameter,
LMD (p) is set to a value indicating a major tone (for example, 0), and LMD (p + LN) is set to a value indicating a minor tone (for example, 1).

In step S14, the major 6th and major 7th notes of the key, which are likely to appear in the minor key, are first and second characteristic sounds CHR1 (p) and CHR2, respectively.
(P). That is, {LTN (l + LN) +9}
mod12 is a major sixth tone (F # in the case of A minor), and {LTN (l + LN) +11} mod12 is a major seven.
It corresponds to the sound of degree (G # in the case of A minor).

In step S15, MKCppmod1
2 (pp = 1 to (the score data number of the last sound of the bar of the bar number MJN)) has the first characteristic sound CHR1 (p) or the second characteristic sound CHR2 (p). Discriminate,
If there is a major key number LTN (p) (p = 0 to LN-
1) and the corresponding mode parameter LMD (p) (p =
0 to LN-1) are deleted. That is, if the characteristic sound of the minor key is included in the first detected section, the major key is deleted. Here, major key number LTN (p)
When (p = 0 to LN-1) is deleted, only the minor key number remains, so the minor key number LTN (p + L
N) and the mode parameter LMD (p + LN) are advanced to LTN (p), LMD (p) (p = 1 to LN-
1). CHR1 (p) or CHR2 (p)
If is not included, do nothing.

In step S16, the key number LTN (p
It is determined whether a plurality of n) remain. LTN here
(Pn) is used to comprehensively display LTN (p) or LTN (p + LN). When only one LTN (pn) remains, the process proceeds to step S20, and when a plurality of LTN (pn) remains, the process proceeds to step S17, and only the key containing the most tones of the melody sounds in the section is selected. That is, the key number LT that maximizes the number of key codes MKCpp for which MKCppmod12 = LTN (pn) holds.
N (pn) and the corresponding mode parameter LMD (p
Only n) is left and the others are deleted (step S1).
7). Here, if there is an LTN (pn) to be deleted, the parameter pn is sequentially advanced. For example, LN =
2 and key numbers LTN (0), LTN (1), LT
When N (2) and LTN (3) remain, LTN
When (0) is deleted, LTN (1), LTN
(2) and LTN (3) are LTN (0) and LTN, respectively.
(1) and LTN (2).

In step S18, the key number LTN (p
n), it is discriminated whether or not a plurality of them remain, and LTN (pn)
If only one remains, go to step S20,
When there are a plurality of remaining tones, the process proceeds to step S19, and only the key containing the largest number of belonging tones is selected and the process proceeds to step S20.
That is, only the key number LTN (pn) and the mode parameter LMD (pn) corresponding to the maximum number of key codes MKCpp for which MKCppmod12 = LTN (pn) +7 are satisfied are left, and the others are deleted (step S19). If there is a key number to be deleted, the parameter pn is advanced as in step S17.

Even if a plurality of key numbers LTN (pn) still remain as a result of the selection in step S19, the process proceeds to step S20, and the key number LTN (0) and the mode parameter LMD (0) are forcibly set, respectively. The keynote TN0 and the mode MD0 are set and stored in the key memory 8.

Next, the key number corresponding to the tonic TN0 and the mode MD0 is set as the current key number CSCLN (step S21), and this routine is ended.

According to the initial key detection subroutine, the degree of coincidence LRA is detected for the first detection section of the melody for key detection.
TE2 (n) is calculated, and key detection is performed based on this degree of coincidence. More specifically, first, the selection candidate key number LSCLN (p) that maximizes the degree of coincidence is obtained (step S12), and the presence or absence of a note characteristic of the minor note (steps S13 to 16), the tonic note, and the group note. Number (step S17
-19), the key of the first detection section is detected.

Next, the key determination subroutine executed in step S6 of FIG. 6 will be described with reference to FIG.

In step S51, the degree of coincidence LRATE2
LRATE in which (n) exceeds a predetermined value R (for example, 90%)
The value n of 2 (n) is the selection candidate key number LSCLN (p), and the number of n is the selection candidate key number LN. Here, the parameter p takes an integer value from 0 to (LN-1).

In step S52, the selection candidate key number LS
The current key number (the key number detected in the preceding detection section) in CLN (p) (p = 0 to LN-1)
It is determined whether or not there is the same key as CSCLN, and if there is the same key (the answer in step S52 is affirmative (YES)), the key number of the detected section this time is the same as the current key number CSCLN (the key is transposed). No)), and this routine is immediately terminated. If the same one as the current key number CSCLN is not in the selection candidate key number LSCLN (p) (the answer to step S52 is negative (NO)), LSCLN (p).
In (p = 0 to LN-1), LRATE1 (n)> L
If there is the same number as the key number n for which RATE2 (n) holds, the selection candidate key number LS for which LSCLN (p) = n
CLN (p) is deleted and the selection candidate key number LN is decreased by the deleted number (step S53). Here, RATE1 (n) is the degree of coincidence in the previous detection section, and the selection candidate key number LSCLN (p) whose degree of coincidence has decreased is removed from the selection candidates in step S53. If there is a key number to be removed, the parameter p is advanced as in the routine of FIG.

In step S54, the selection candidate key number LS
It is determined whether or not CLN (p) remains, and if it does not remain (the answer in step S54 is negative (NO)), it is determined that there is no modulation, and this routine is immediately terminated. When the selection candidate key number LSCLN (p) remains (the answer in step S54 is affirmative (YES)), LSCLN (p) (p = 0 to LN- is used by using the priority table of FIG.
Among the keys corresponding to 1), the key with the highest possibility of being transposed (priority order) is obtained. The first key of the priority table of FIG. 3 is the key (TN) corresponding to the current key number CSCLN.
j-1, MDj-1).

In step S56, it is determined whether or not there is a corresponding one in the priority table, and if there is no corresponding one (the answer in step S56 is negative (NO)), it is determined immediately that no modulation is performed. This routine ends. When there is a corresponding one (the answer in step S56 is affirmative (YES)), the bar number i of the first bar of the detection section which is the target of the current detection, the tonic of the corresponding key and the mode of the priority table are respectively set. The bar number MEj, the tonic TNj, and the mode MDj stored in the key memory are set, and the key number corresponding to TNj and MDj is set as the current key number CSCLN (step S57). Then, the key data number j is incremented by 1 (step S5
8), the current calculated value of the degree of coincidence LRATE2 (n) (n = 0
~ 11) as LRATE1 (n) (step S5
9), this routine is ended.

According to the key determination subroutine of FIG. 9, a key whose degree of coincidence LRATE2 (n) with the in-tone exceeds a predetermined value R is selected as a candidate key for selection, and some of these keys are the same as the previous key. If it is not transposed, it is determined. If there is no match, the one whose degree of coincidence decreases is removed from the candidates, and if there are still candidates, it is determined that the key with the highest possibility of modulation in the priority table has been transposed. ..

Next, the above-mentioned key detection method will be described with reference to FIGS.

In the example of FIG. 11, the section length MEN of the detection section
= 4, the shift length SME = 1, and the degree of coincidence is shown only in the upper two. Further, FIG. 12 shows the transition of the degree of coincidence.

First, regarding the first detection section (first to fourth measures), key detection is performed by the initial key detection subroutine of FIG. The degree of coincidence LRATE1 (n) is the maximum when LRATE1 corresponding to C major or A minor
Since it is only (0), the selection candidate key number LSCLN
(0) = 0 and the selection candidate key number LN = 1.

Since there is no major 6th or major 7th note (F #, G #) from the A minor tonic A in the first detection section, C major is not deleted in step S15 of FIG. In the first detection section, the tonic C of C major is 4
Since there are three main tones A of the A minor, the A minor is removed from the candidates in step S17 of FIG. 7, and the key of the first detection section is detected as C major.

In the second detection section (second to fifth measures), if the predetermined value R is 90%, the selection candidate key number LSCLN
(0) = 0, LSCLN (1) = 5, selection candidate key number LN
= 2. Here, since the current key number CSCLN = 0, the answer in step S52 in FIG. 9 is affirmative (YES).
Therefore, it is determined that the key is not transposed.

Third and fourth detection sections (third to sixth measures,
The fourth to seventh measures) are also determined in the same manner as the second detection section.

In the fifth detection section (5th to 8th measures), the selection candidate key number is only LSCLN (0) = 5 (F), and the answer to step S52 in FIG. 8 is negative (NO).
Since the matching degree of LSCLN (0) = 5 is the same as the previous time (97%), it is not deleted in step S53 of FIG.
Since there is F major as the key that is most easily transposed from the previous major C major in the priority table of No. 2, the answer in step S56 is affirmative (YES), and it is determined that the major major has been transposed. Therefore, the fifth measure which is the first measure of the fifth detection section
It is detected that the measure has changed to F major.

As described above, according to the present embodiment, the detection section is set to be relatively long (for example, 4 bars), and the shift length is set to be short (for example, 1 bar) to include a portion to be transposed. It is possible to accurately detect the melody key and specify the transposed point with high resolution.

In calculating the degree of coincidence with the in-tone,
Weighting may be performed according to whether or not the note data is a strong beat.

The supply of the musical score data is not limited to the external storage device such as a floppy disk or the like, and may be performed by the performer in real time or in step writing.

The musical score data is not limited to a melody, but may be an accompaniment sound or the like.

FIG. 10 shows step SM of the main routine.
6 is a flowchart of a subroutine of automatic arrangement processing shown by 2.

First, in step S61, the user receives an input of an arrangement style such as rock, disco, jazz, tango, or rumba. Next, step S62.
Then, in the key detection routine, TNj and MDj stored together with the bar number MEj are searched in order from the beginning, and a chord is recorded for each bar from the beginning to the bar immediately preceding the bar having the key change for the first time. Give. When adding chords, the six major chords (I, IIm, IIIm, IV, V, VI) in the major tones that are often used for the tones are based on the tones of the section.
m), major 8 chords in the case of minor tone (Im, III, IV, IV
m, V, Vm, VI, VII) is calculated by calculating the degree of coincidence between the note names of the constituent notes and the note names in the melody and selecting the highest one. If they have the same degree of coincidence, they are decided with appropriate priority. According to music theory, it is desirable that the priority order be varied depending on the style of arrangement.

The chords added to each bar in this way are stored in the chord memory 11 together with the bar numbers.

Next, a section from the bar position at which there is a change in key for the first time to the bar immediately before the bar at which there is a next change is set as a section, and a chord is similarly added to each bar, and a chord memory is stored together with the bar number. I will remember it in 11. After that, the same operation is performed until the end of the melody data.

In step S63, the rhythm pattern number in the pattern memory 12 corresponding to the arrangement style input in step S61 is stored in the accompaniment performance data memory 10 in correspondence with the measure. There are many variations in the same arrangement style, so choose from them at random.

Next, in steps S64 and S65, in each step, the base pattern or backing pattern in the pattern memory 12 corresponding to the style is repeatedly read, converted into a pitch corresponding to a chord, and converted into note data. It is stored in the accompaniment performance data memory 10.
Since these patterns also have various variations with respect to the same arrangement style, they are randomly selected from them and converted into note data and stored.

Finally, in step S66, a harmony tone which is 3 degrees below or 3 degrees below or 4 degrees below the melody is selected based on the key data and the chord, and is similarly stored as accompaniment performance data memory as note data. Store in 10.

As described above, automatic arrangement is performed based on the data obtained by the key detection subroutine.
After that, the process returns to the main routine and all the arranged part data of the accompaniment performance is transferred to the floppy disk.

It should be noted that a subroutine for viewing after editing may be provided. In this case, the note data stored in the accompaniment performance data memory 10 and the melody memory 7 is output from the output interface 13 shown in FIG. 1, and for the rhythm pattern stored as the pattern sequence data, The pattern memory 12 may be referred to so as to be converted into musical note data and outputted, and a musical tone may be generated and viewed by a sound source device connected to the output interface 13.

Further, a correction routine may be provided for replacing the selection pattern with another candidate for a portion which the viewer does not like after viewing.

Alternatively, the arranged floppy disk may be set in an automatic performance device of the same format and automatically played.

[0074]

As described above in detail, according to the first aspect of the invention, even if the music data contains a modulation, the modulated key and the modulation position are detected, and the key and the position are detected. Since the music is arranged on the basis of the arrangement, it is possible to arrange the music data including the modulation without any inconvenience.

According to the second aspect of the present invention, the key detection is performed based on the degree of coincidence between the note data and the intone, and the predetermined section as a unit for calculating the degree of coincidence has an overlapping portion with an adjacent section. Since it is specified as described above, it is possible to accurately detect the key of the melody including the part to be transposed, and to specify the transposition location with high resolution.

[Brief description of drawings]

FIG. 1 is a block diagram showing an overall configuration of a key detection device according to an embodiment of the present invention.

FIG. 2 is a diagram showing a format of data stored in a melody memory and a tone memory.

FIG. 3 is a diagram showing a priority order table used to determine a key that has been transposed.

FIG. 4 is a diagram showing a note number data structure of a key number (n) and a key note.

FIG. 5 is a flowchart of a main routine of key detection processing and automatic music arrangement processing.

FIG. 6 is a flowchart of a subroutine of key detection processing.

FIG. 7 is a flowchart of a subroutine for performing initial tone detection.

FIG. 8 is a flowchart of a subroutine for detecting the degree of coincidence between note data and intones.

FIG. 9 is a flowchart of a subroutine for determining whether or not a transposition is performed during a melody.

FIG. 10 is a flowchart of a subroutine of automatic arrangement processing.

FIG. 11 is a diagram showing an example of musical score data.

FIG. 12 is a diagram showing the transition of the degree of coincidence in the data example of FIG. 9.

[Explanation of symbols]

 1 floppy disk device 2 disk control device 3 central processing unit (CPU) 7 melody memory 8 tone memory 9 priority table 10 accompaniment performance data memory 11 chord memory 12 pattern memory

Claims (2)

[Claims] 1. A music data storage means for storing music data, a key detection means for analyzing the music data and detecting a key of the music data and a key position where the key is transposed, the key detection means. A key storage means for storing key data representing, wherein when the music data is transposed, key data corresponding to sequentially changing keys is stored in correspondence with the position of the music data, On the basis of the key data stored in the key storage means, when the key is transposed, the key that sequentially changes is played together with the music data also based on the position of the corresponding music data stored. An automatic arrangement device including a music data generation unit that generates other music data to be reproduced. 2. Music data supply means for supplying music data including note data, section specifying means for sequentially specifying a predetermined section of the music data, and note names and notes of note data in the specified predetermined section. A coincidence degree calculating means for obtaining a degree of coincidence with a note name of an internal scale note, a specifying means for sequentially specifying the predetermined section so as to have an overlapping portion with an adjacent section, and a forward calculation for the sequentially specified section. And a modulation means for detecting modulation according to the degree of coincidence.