JP4132239B2 - Sound editing system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a sound editing system for PCM sound signals mediated by MIDI editing.
[0002]
[Prior art]
Conventionally, multi-track waveform editing based on the PCM (Pulse Code Modulation) method has been most widely used for sound signal editing. However, due to the spread of computer music composition methods in recent years, it is based on the MIDI (Musical Instrument Digital Interface) standard. Code data (hereinafter referred to as MIDI data) has also been actively used. A MIDI track can be assigned to a part of tracks of the editing device (some devices have a maximum of 256 tracks), and PCM and MIDI can be edited on the same time axis.
[0003]
However, even if the editing operation is added to the PCM track and the MIDI track in a broadcast manner, each is edited independently.
(1) Abundant score editing functions are limited to MIDI tracks.
(2) Mutual boarding (cutting editing) is not possible between PCM and MIDI tracks.
There are functional restrictions such as.
[0004]
Conversion from MIDI data to PCM data can be originally performed by sound source reproduction. If a means for converting PCM data to MIDI data is provided, the score editing function can be used for all tracks. For example, although conversion is possible according to Japanese Patent Application Nos. 9-67467 and 9-249635, there is a problem in terms of quality when MIDI data is edited and returned to PCM data again.
[0005]
[Problems to be solved by the invention]
The present invention has been made in view of such problems, and an object of the present invention is to provide an acoustic editing system that can maintain the editing quality of PCM data and can use the editing function of MIDI data. .
[0006]
[Means for Solving the Problems]
In order to achieve the above-described object, the present invention provides a score conversion means for converting a source sound signal recorded in a waveform format into score-format score data, and interactively edits each note constituting the score data. A score editing means for generating edited score data, and a difference score data between the edited score data and the score data, which is a change of the score data edited with respect to the notes by the score editing means. the acoustic command translation means for translating the acoustic signal commands performed on the signal segment corresponding to the notes on the source audio signal, the source sound based on the acoustic signal commands translated by the acoustic command translation means An acoustic editing system comprising: an acoustic signal editing unit that edits the signal section on the signal .
[0007]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a schematic configuration diagram of an audio editing system 1 mediated by MIDI editing according to an embodiment of the present invention. As shown in FIG. 1, the sound editing system 1 includes a MIDI conversion unit 3, a MIDI editing unit 5, a MIDI data editing command translating unit 7, a PCM sound signal batch editing unit 11 using a PCM sound editing command 9, and the like.
[0008]
The MIDI conversion unit 3 converts a general PCM-encoded source sound signal (hereinafter referred to as source PCM data) 15 into MIDI data 21 in the MIDI format. This conversion is performed based on the methods of Japanese Patent Application Nos. 9-67467 and 9-249635. The MIDI editing unit 5 edits the converted MIDI data 21 and creates the edited MIDI data 23.
[0009]
The editing command translation unit 7 generates a PCM sound editing command 9 from the difference MIDI data 25 between the edited MIDI data 23 and the converted MIDI data 21, which is a change in the MIDI data edited by the MIDI editing unit 5. . The PCM sound signal batch editing unit 11 performs sound signal editing on the source PCM data 15 using the generated PCM sound editing commands 9 group, and creates post-editing PCM data 17.
[0010]
Next, the relationship between editing by the MIDI editing unit 5 and the PCM sound editing command 9 will be described. FIG. 2 is a diagram showing the relationship between the editing by the MIDI editing unit 5 and the PCM sound editing command 9.
[0011]
First, editing by the MIDI editing unit 5 will be described. MIDI data is basically data describing information about individual notes (notes). As shown in FIG. 2, the MIDI editing unit 5 performs note deletion 31, addition 33, and attribute change 35. The attribute change 35 is a change of the sound start time (note-in / delta time) 36, a sound end time (note-off delta time) 37 (change of the sound length), and the pitch (note number). ) 38, and sound intensity (velocity) 39.
[0012]
Here, four pieces of information, which are a sound start time 36, a sound length 37, a sound pitch 38, and a sound intensity 39, which are the basis of the MIDI data editing process, will be described. In order to display the MIDI data, it is suitable to display in the form of a score using a code according to a note. Japanese Patent Application No. Hei 9-67468 defines a note-like code that is a trapezoidal or rectangular note graphic, two-dimensional with the horizontal axis representing the time / length axis and the vertical axis representing the pitch / sound intensity axis. Place a note graphic on the coordinates. Thereby, display in the form of a score is performed. This will be described using this display method. FIG. 9 is a diagram showing a MIDI data editing method using note graphics.
[0013]
As shown in FIG. 9, the MIDI editing unit 5 displays quantitatively based on the four pieces of information of the sound start time t1, the sound length d1, the sound pitch n1, and the sound intensity v1. And edit interactively. For example, the editing from the note graphic Z1 to the note graphic Z2 is a change of the start time t1 for horizontally moving the note graphic and a change of the pitch n1 for vertically moving the note graphic (Dracking 27). The editing from the note graphic Z1 to the note graphic Z3 is a change of the sound length d1 for performing horizontal expansion / contraction of the musical note graphic and a change of the sound intensity v1 for performing vertical expansion / contraction of the musical note graphic (resizing 29).
[0014]
The change range is 0 to 128 ^{3 for} the sound start time change 36, 0 to 127 for the sound length change 37, 0 to 128 ^{3 for} the sound pitch change 38, and 0 to 128 for the sound intensity change 39. There are 127 stages.
[0015]
Next, the relationship between the MIDI editing unit 5 and the PCM sound editing command 9 will be described. Editing a note in the MIDI editing unit 5 corresponds to performing equivalent editing on a predetermined signal section in the PCM sound editing command 9. The PCM sound editing command 9 automatically deletes, adds, and moves, extends, expands, and contracts the waveform of the signal section with respect to the horizontal signal time axis and the vertical signal intensity axis.
[0016]
As shown in FIG. 2, note deletion 31 corresponds to section signal deletion 41, and note addition 33 corresponds to section signal addition 43. The change of the sound start time 36, the sound length 37, the sound pitch 38, the sound intensity 39, etc. in the note attribute change 35 is performed by changing the section position 46 and the section length of the section signal in the section signal attribute change 45. This corresponds to a change in the length 47, the signal pitch 48, the signal amplitude 49 and the like.
[0017]
Next, the change of the section length 47 and the change of the signal pitch 48 in the PCM sound editing command 9 will be described in detail. FIG. 3 is a diagram illustrating a change in the length 47 of the section. FIG. 4 is a diagram illustrating the change of the signal pitch 48.
[0018]
As shown in FIG. 3, the section extension 51 divides the waveform of the current PCM signal section 50 at a constant time interval and extends it to a desired section with an equal interval. Either forward or backward signals are copied at each interval, but if the front is extended backward, the stretched tip is continuous with the rear, but is generally discontinuous at the stretched source. Although the occurrence of discontinuity cannot be avoided at the connection portion of this waveform, it is desirable to shift the connection position so as to reduce the amplitude gap at the discontinuity point and bring the discontinuity point as close to 0 level as possible.
[0019]
In the example of FIG. 3, the signal period 50 of about 5 cycles is extended to the signal period of about 8 cycles. First, it is divided into three by two periods from the rear (A1, A2, A3), and one rear period (A1, C1, C2) of each divided period section is copied (52).
[0020]
The section shortening 53 divides the waveform of the current PCM signal section 50 at a constant time interval and shortens the waveform to a desired section while deleting signals equally. The connecting points before and after the deleted section are generally discontinuous. Similarly to the section extension 51, it is desirable to shift the section to be deleted back and forth and bring discontinuities as close to the 0th level as possible.
[0021]
In the example of FIG. 3, the signal period 50 of about 5 cycles is shortened to the signal period of about 3 cycles. First, it is divided into two periods from the rear (B1, B2, B3), and the rear one period (D1, D2) of the divided two period sections (B2, B3) is deleted (54).
[0022]
Next, as shown in FIG. 4, the signal pitch increase 55 reduces the current PCM signal section 50 in the signal time axis direction, which is the horizontal axis, so as to have the same section length L as the current PCM signal section 50. The section is extended in the same manner as the section extension 51 described above (56). The signal pitch decrease 55 enlarges the current PCM signal section 50 in the signal time axis direction, and shortens the section in the same manner as the section shortening 53 described above so as to have the same section length L as the current PCM signal section 50. (58).
[0023]
Next, the process of FIG. 1 will be described using a track. FIG. 5 is an explanatory diagram for explaining the processing of FIG. 1 using a track.
[0024]
When copying between PCM tracks 13 and MIDI tracks 19, there is no particular processing other than the conversion of attribute information. When copying between PCM tracks 13, the sampling frequency of attribute information and the signal quantization bit length are matched. When copying between MIDI tracks 19, time units such as metronome and time signature are matched.
[0025]
In FIG. 5, the PCM track 61 holds source PCM data 15. The MIDI conversion unit 3 performs MIDI conversion on the source PCM data 15 and copies it to the MIDI track 62. At this time, link information 65 having the track ID and section information of the source PCM track 61 is added to the MIDI track 62 .
[0026]
Next, the converted MIDI data 21 held in the MIDI track 62 by the MIDI editing unit 5 is edited to create the edited MIDI data 23, and the created edited MIDI data 23 is held in the MIDI track 63. The PCM sound editing command 9 is generated from the difference data 25 using the converted MIDI data 21 held in the MIDI track 62 and the edited MIDI data 23 held in the MIDI track 63.
[0027]
Then, the PCM sound signal batch editing unit 11 edits the source PCM data 15 held in the PCM track 61 by using the PCM sound editing command 9 to obtain the edited PCM data 17, and the edited PCM data. 17 is held on the PCM track 64. At this time, the edited PCM data 17 is held in the PCM track 64 indicated by the link information 65 given to the MIDI track 62 that is the basis of the PCM sound editing command 9 and the predetermined section.
[0028]
As described above, in the present embodiment, the PCM-encoded source acoustic signal (source PCM data) 15 is converted into the MIDI data 21 by the MIDI conversion unit 3, and the edit command translation unit is changed from the change in the MIDI edited MIDI data. 7, the PCM sound editing command 9 is generated, and the PCM sound signal batch editing unit 11 edits the command 9 group generated for the source PCM data 15. As a result, the PCM sound data can be edited in the same way as the MIDI data, and a MIDI editing environment can be provided for non-musical sound materials.
Further, mutual loading (cutting editing) can be performed between the MIDI track 19 and the PCM track 13, and deterioration of the sound quality of the source PCM data 15 can be suppressed.
[0029]
Next, multitrack MIDI editing and PCM sound editing will be described. FIG. 6 is a diagram showing multitrack MIDI editing and PCM sound editing.
[0030]
As shown in FIG. 6, when the source acoustic signal is composed of a multi-track source acoustic signal L and a source acoustic signal R, any one track is selected, and a MIDI conversion unit for the source acoustic signal is selected. MIDI conversion is performed by 3a. For example, the source acoustic signal L is selected and MIDI conversion is performed on the source acoustic signal L.
[0031]
When converting one source sound signal into multi-track MIDI data, the MIDI editing unit 5a performs MIDI editing on any one of the MIDI tracks. For example, MIDI conversion is performed on the source sound signal L to generate three MIDI tracks-accompaniment melody track 95, vocal melody track 96, and vocal formant track 97. Then, MIDI editing is performed with reference to the MIDI data 98 which is another music melody.
[0032]
Further, a PCM sound editing command is generated from the edited MIDI data by the editing command translating unit 7a, and the source sound signal is edited by the source PCM batch editing unit 11a. In this case, the same editing process is performed for all the source audio signals of the multitrack. For example, the editing command translation unit 7a generates the editing L and editing R of the PCM sound editing command from the edited MIDI data for the accompaniment melody track 95, and the editing L and editing R and the editing sound L and editing R of the source sound signal L by the PCM sound signal batch editing unit 11a Edit R is performed on the source acoustic signal R. Thereby, the edit sound signal L and the edit sound signal R are created. (Edit L and Edit R have the same editing content.)
[0033]
Next, a method for creating a replacement song by synthesizing two songs will be described. FIG. 7 is a diagram illustrating a method of creating a replacement song by combining two songs.
[0034]
As shown in FIG. 7, the source PCM data 71 of the song A and the source PCM data 75 of the song B are converted by the multi-track MIDI conversion unit 3b, respectively. A conversion MIDI track A72 and a conversion MIDI track B76, which are melody tracks, are selected from the multitrack MIDI data. For example, editing is performed by the MIDI editing unit 5b so that the melody track of the song A is replaced with the melody track of the song B. That is, difference data 73 of (converted MIDI track B-converted MIDI track A) is generated, a PCM sound editing command is generated from the difference data 73 by the edit command translating unit, and editing is performed by song PC A by the PCM sound signal batch editing unit 11b. To the source PCM data 71. Thereby, a song 74 having the lyrics of song A with the melody of song B is created.
[0035]
Similarly, the source PCM data 75 of the song B is edited by the PCM sound signal batch editing unit 11b using the difference data 77 of (converted MIDI track A-converted MIDI track B), and the song B's melody of the song B is edited. A song 78 with lyrics is created.
[0036]
Next, a specific example of generating a replacement song will be described. FIG. 8 shows a specific example of the creation of a singing song by composing Yamada Kogo's “Red Dragonfly” and “Kono Michi”.
[0037]
This parody changes only the pitch of the melody and associates the rhythm of the lyric with phrases that seem to have good engagement with the length of the sound. As shown in FIG. 8, the MIDI data of the melody tracks of the singing voices of “Red Dragonfly” and “Kono Michi” are extracted, and the lyrics 81 of the “Red Dragonfly”, the note 82, the pitch (note number) 83, the length of the sound. MIDI editing is performed on the information of the lyrics 84, the note 92, the pitch 92 (note number) 93, and the length 94 of the sound. The notes 82 and 92 correspond to “do”, “le”, “mi”, and the like.
[0038]
The lyrics 81 and 91 and the sound lengths 84 and 94 of the lyrics of “Red Dragonfly” and “Kono Michi” are compared. For example, the rhythm of Lyrics 81-3 (Akatonbo) of “Akatonbo”, the rhythm of Lyrics 91-1 (Kono Michiha) of “Kono Michi”, and the length of the sound of “Red Dragonfly” 84-3 (1 1 2 2 4) and the length of the sound of this road 94-1 (1 11 2 3) are quite similar, so the third phrase of “Red Dragonfly” and the first of “this road” The phrase is matched. Similarly, the first phrase of “Red Dragonfly” is associated with the second phrase of “this road”, the second phrase of “Red Dragonfly”, and the third phrase of “this road”.
[0039]
Next, differential data 96 of the pitch (note number) between the associated phrases is obtained. That is, the pitch 96-1 in the first phrase of the generated singing song is (pitch 93-1 of "this road")-(pitch 83-3 of "red dragonfly") The difference data (−13 + 1 + 3 + 3 + 1) is obtained. Similarly, the pitch of the sound 96-2 is the difference data (+ 15 + 6 + 4 + 3) which is (the pitch of the sound of "this road" 93-2)-(the pitch of the sound of "red dragonfly" 83-1). The pitch of the sound 96-3 is the difference data (−1 −2 −19 −) (the pitch of the sound of “this road” 93-3) − (the pitch of the sound of “red dragonfly” 83-2). 6-6) is obtained.
[0040]
Also, by replacing the lyrics between the associated phrases, the lyrics 95-1 in the first phrase of the generated percussion is changed to the lyrics 81-3 of “Red Dragonfly”. Similarly, the lyrics 95-2 is changed to “red dragonfly” lyrics 81-1 and the lyrics 95-3 is changed to “red dragonfly” lyrics 81-2.
[0041]
In this way, a percussion song is generated in which “red dragonfly” and “this road” are synthesized by changing only the pitch of the notes and associating the phrase in which the length of the sound and the rhythm of the lyrics mesh with each other.
[0042]
Therefore, in the case of the PCM data 15 in which the source sound signal is edited, if a portion corresponding to the melody of the BGM or accompaniment part is extracted by the MIDI conversion unit 3 and changed to another melody by the MIDI editing unit 5, a song is replaced Can be created, and melodies can be added to the readings of poetry.
[0043]
That is, in the present embodiment, when MIDI data that does not correspond to the source sound signal is separately provided, the MIDI editing unit 5 converts a part of the MIDI data to be edited into a part of the separately provided MIDI data. Editing can be done in a similar way. Note that separately provided MIDI data does not have to be converted from another source sound signal.
[0044]
【The invention's effect】
As described above in detail, according to the present invention, it is possible to provide an acoustic editing system that can maintain the editing quality of PCM data and can use the editing function of MIDI data.
[Brief description of the drawings]
FIG. 1 is a schematic configuration diagram of an audio editing system 1 mediated by MIDI editing according to an embodiment of the present invention. FIG. 2 is a diagram showing a relationship between editing by a MIDI editing unit 5 and PCM audio editing commands 9. 3] Diagram showing change in section length 47 [FIG. 4] Diagram showing change in signal pitch 48 [FIG. 5] Explanatory diagram explaining processing in FIG. 1 using tracks [FIG. 6] Multi-track MIDI editing Fig. 7 shows the editing of PCM sound. Fig. 7 shows how to create a replacement song by combining two songs. Fig. 8 shows a specific example of the creation of a replacement song by Koda Yamada's combination of "Red Dragonfly" and "Kono Michi." 9] Diagram showing how MIDI data is edited using note graphics [Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ......... Sound editing system 3 ......... MIDI conversion part 5 ......... MIDI editing part 7 ......... Editing command translation part 11 ......... PCM sound signal batch editing part 13 ......... PCM track 15 ......... Source PCM data 17 ......... Edited PCM data 19 ......... MIDI track 21 ... ... Converted MIDI data 23 ... ... Edited MIDI data 25 ... ... Difference MIDI data 31 ... ... Delete note 33 ... ... Note Added 35, 45 ......... Change attribute 36 ......... Start time of sound (note on delta time)
37 ……… End time of sound (note-off delta time)
38 ......... Pitch (note number)
39 ……… Sound intensity (velocity)
41 ......... Section signal deletion 43 ......... Section signal addition 46 ......... Section position 47 ... ... Section length 48 ... ... Signal pitch 49 ... ... Signal amplitude 65 ... ... Link information

Claims (6)

A score conversion means for converting a source acoustic signal recorded in a waveform format into score format score data;
A score editing means for interactively editing each note constituting the score data and generating score data after editing,
From the difference score data between the edited score data and the score data, which is a change of the score data edited with respect to the note by the score editing means, to the signal section corresponding to the note on the source acoustic signal Acoustic command translation means for translating into acoustic signal commands performed on
Acoustic signal editing means for editing the signal section on the source acoustic signal based on the acoustic signal command translated by the acoustic command translation means;
An acoustic editing system comprising: When the score conversion means converts to a score data consisting of a plurality of tracks that overlap in time with respect to one source acoustic signal,
2. The sound editing system according to claim 1, wherein the musical score editing means targets any one of the representative tracks.   The score editing means operates in an interactive manner on a score that quantitatively displays four pieces of information of time position and length of each note and pitch and strength of each note. The sound editing system according to 1.   2. The sound editing system according to claim 1, wherein the source sound signal is encoded in a PCM format, and the musical score data is encoded in a MIDI format.   2. The musical score editing unit according to claim 1, wherein the musical score editing unit performs any one of deletion of the note, addition of the note, and change of time position, length, height, and strength attribute of the note. Sound editing system.   2. The sound according to claim 1, wherein the acoustic signal command is one of deletion of a section signal, addition of a section signal, section position of a section signal, section length, signal pitch, and signal amplitude attribute change. Editing system.

Images