JP4274272B2 - Arpeggio performance device - Google Patents

Description

  The present invention relates to an arpeggio performance device that automatically performs an arpeggio performance. In particular, the present invention relates to an arpeggio performance device that can easily perform more realistic and diverse arpeggio performances using a phrase waveform that records actual arpeggio performances.

Conventionally, a plurality of pitch information generated in response to the operation of a performance operator or a plurality of pitch information based on a specified chord, for example, a predetermined arpeggio (distributed chord) pattern such as up, down, up-down, etc. In addition, an arpeggio performance apparatus (also called an arpeggiator) that automatically performs arpeggio performance by controlling the sound generation timing is known. In general, it is very difficult to perform an arpeggio performance by manual (hand-playing) performance in which the performer himself / herself actually operates the performance operator. For example, when performing an arpeggio using an acoustic piano or guitar, it is required to quickly operate multiple performance controls (keys, strings, etc.) in a predetermined order, especially for beginners. It is difficult to perform the operation accurately. However, if the arpeggiator is used, the pitch information generated in response to the operation is applied to a predetermined arpeggio pattern and the musical sound is reproduced simply by operating a plurality of performance operators almost simultaneously. You can perform arpeggios simply by performing simple performance operations.
Recently, a phrase recorded by a performer using a natural instrument (live instrument) is actually recorded manually, and the tempo and pitch (pitch) of the recorded phrase are changed. There is known a waveform data reproducing apparatus called a so-called phrase sampler. As an example of such a sampler, there is an apparatus described in Patent Document 1 shown below. In the apparatus described in Patent Document 1, instrument sounds from a manually-played natural instrument are recorded synchronously in a phrase unit while reproducing predetermined performance data. Then, the recorded series of phrase waveforms are divided and recorded into partial waveforms for each sound based on the performance data, and the recorded partial waveforms are individually specified to generate sound generation timing and pitch (pitch). The music can be played while changing the above.
JP 2002-229567 A

However, in the conventional arpeggiator as described above, a sound source such as a PCM sound source or an FM sound source, for example, according to a plurality of pitch information generated according to the operation of the performance operator or a plurality of pitch information based on a specified chord. Is driven by a predetermined arpeggio pattern, and the arpeggio performance is generated, so it is an arpeggio performance in which each sound is simply arranged according to the arpeggio pattern, and there is no connection between sound and sound. There was a problem that it lacked realism compared to an arpeggio performed using natural instruments.
On the other hand, with a conventional sampler as described above, it is possible to record an arpeggio performance such as an arpeggio performed with a natural instrument and reproduce the phrase waveform data to express the arpeggio performance. The musical sound generated in this way is more realistic than the musical sound generated by the arpeggiator. The sampler can also be played with the tempo and pitch changed. However, these samplers are inconvenient if you do not record a phrase performance corresponding to each arpeggio pattern in advance, and you cannot perform arpeggio performances with different arpeggio patterns according to the pattern specification. Arpeggio performances with different chords cannot be performed accordingly, and in that case, it is troublesome because it is necessary to instruct the pitch change for each note.

  The present invention has been made in view of the above points, and it is possible to easily perform a more realistic and diverse arpeggio performance using phrase waveform data in which an arpeggio performance actually performed in advance is recorded. The device is to be provided.

The arpeggio performance device according to claim 1 of the present invention is a phrase waveform corresponding to the arpeggio pattern indicated by the pattern ID, respectively, a first storage means for storing a plurality of phrase waveforms divided into a plurality of partial waveforms, Corresponding to each of the plurality of phrase waveforms, second storage means for storing characteristic information such as pitches and sound generation timings related to the partial waveforms obtained by dividing the phrase waveform, and pattern IDs and codes relating to the arpeggio performance to be reproduced , A means for generating a normative arpeggio pattern corresponding to the designated pattern ID and code, and a phrase waveform corresponding to the designated pattern ID stored in the first storage means Each of the partial waveforms of the characteristic information stored in the second storage means and the generated normative arpeggio pattern And modified based on the bets, comprising a reproducing means for generating a performance waveform data of the arpeggio.
According to the present invention, the user is allowed to specify the pattern ID and chord related to the arpeggio performance to be reproduced, the reference arpeggio pattern corresponding to the designated pattern ID and chord is generated, and the pattern ID corresponding to the designated pattern ID is generated. Each partial waveform of the stored phrase waveform is corrected based on the feature information stored in the second storage means and the generated standard arpeggio pattern to generate performance waveform data of the arpeggio performance. Performance waveform data of an arpeggio performance with a chord different from that of the phrase waveform can be reproduced from the phrase waveform that is present.

The arpeggio performance device according to claim 2 further comprises means for allowing the user to specify a tempo related to the arpeggio performance to be played back, and means for generating a tempo clock of the specified tempo. The reproduction means corrects the reproduction timing and the sound length of each partial waveform based on the generated tempo clock.
According to this, performance waveform data of an arpeggio performance having a tempo different from the tempo of the phrase waveform can be reproduced from the stored phrase waveform.

The arpeggio performance device according to claim 3 is the arpeggio performance device according to claim 1, wherein a plurality of phrase waveforms corresponding to one pattern ID are stored in the first storage means, and also stored in the first storage means. Each phrase waveform is associated with a unique phrase ID, and the reproduction means corresponds to one pattern ID and shows a phrase ID for each range related to the root sound or type of the chord. A phrase ID corresponding to the specified pattern ID and corresponding to the range including the specified code is obtained based on the third storage means storing the data and the map data of the third storage means And means for reading out a phrase waveform corresponding to the acquired phrase ID from the first storage means.
According to this, it is possible to change the phrase waveform used for reproducing the performance waveform data of the arpeggio performance according to the designated chord range, and to generate performance waveform data of a more natural arpeggio performance.
According to a fourth aspect of the present invention, in the second aspect, the first storage means stores a plurality of phrase waveforms corresponding to one pattern ID, and the first storage means stores the plurality of phrase waveforms. Each phrase waveform is associated with a unique phrase ID, and the playback means corresponds to one pattern ID and maps data indicating the phrase ID for each range related to the chord and the tempo. Corresponding to the designated pattern ID and corresponding to the range including the designated code and the designated tempo based on the stored third storage means and the map data of the third storage means Means for acquiring a phrase ID; and means for reading out a phrase waveform corresponding to the acquired phrase ID from the first storage means. To.
According to this, it is possible to change the phrase waveform used to reproduce the performance waveform data of the arpeggio performance according to the specified chord and the specified tempo range, and generate more natural arpeggio performance waveform data. it can.

  The present invention can be constructed and implemented not only as a device invention but also as a method invention. Further, the present invention can be implemented in the form of a program of a processor such as a computer or a DSP, or can be implemented in the form of a storage medium storing such a program.

Embodiments of the present invention will be described below in detail with reference to the accompanying drawings.
FIG. 1 is a hardware configuration block diagram showing an embodiment of the overall configuration of an arpeggio performance apparatus according to the present invention. The arpeggio performance device (arpegator) shown in this embodiment is controlled by a microcomputer comprising a microprocessor unit (CPU) 1, a read only memory (ROM) 2, and a random access memory (RAM) 3. That is, the CPU 1 is a central processing unit that controls recording and playback operations in the arpeggiator by executing various programs. The timer 1A is a timer that measures an elapsed time during operation and generates an interrupt time in timer interrupt processing (interrupt processing), and is used for time management and the like. The ROM 2 stores various programs executed by the CPU 1 and various data. The RAM 3 is used as a working memory that temporarily stores various data generated when the CPU 1 executes a predetermined program, or as a memory that stores a currently executed program and related data. A predetermined address area of the RAM 3 is assigned to each function and used as a register, flag, table, memory, or the like.

  The display 4 is a display such as a liquid crystal display panel (LCD) or a CRT, for example, and is used for recording an arpeggio performance (actual performance) using a natural musical instrument by the performer himself or reproducing phrase waveform data described later. Information is displayed or the control state of the CPU 1 is displayed. The operation element 5 is an operation element for performing various designations and instructions when recording the actual performance or reproducing the waveform data. For example, the operation element 5 is an automatic performance based on pattern performance data (see FIG. 3 to be described later). Arpeggio performance) for starting / stopping recording, recording switch for instructing recording start / stop of actual performance, pattern designating switch for designating a predetermined arpeggio pattern, chord designating switch for designating a chord, arpeggio performance There is a map selection switch for selecting assigned map data (see FIG. 7 described later). Of course, in addition to these, a numeric keypad for inputting numeric data, a keyboard for inputting character data, and a predetermined pointing device displayed on the display 4 are used for selecting, setting, and controlling pitches, tones, effects, and the like. Various operators such as a mouse used for operating the mouse may be included. Further, it may include a performance operator such as a keyboard provided with a plurality of keys for selecting the pitch of the musical sound. The performer operates these controls 5 while referring to various information displayed on the display 4 to confirm phrase waveform data (see FIG. 5 to be described later) in which, for example, an actual arpeggio performance is recorded. It is possible to easily perform processing, edit map data that has already been created, and select map data to be used.

  The tone generator 6 can generate musical tone signals simultaneously on a plurality of channels, and inputs various performance information according to data and pattern performance data (see FIG. 3) that is automatically played during recording, which is given via the address bus 1D. A musical tone signal is generated based on the performance information. Such a sound source 6 may employ any of various tone synthesis methods such as FM, PCM, physical model, formant synthesis, etc., may be configured by dedicated hardware, or configured by software processing by the CPU 1. May be. A musical sound signal generated from the sound source 6 is generated from the sound system 9 via the mixer 7 and the DAC 8. The mixer 7 is a mixer that mixes the musical tone signal generated by the sound source 6 and the musical tone signal reproduced by the reproduction circuit 14 to be described later and sends it to the DAC 8. The DAC 8 converts the digital musical tone signal supplied from the mixer 7 into an analog signal. It is a digital-analog signal converter that converts a musical sound signal and sends it to the sound system 9. The sound system 9 includes an amplifier and a speaker for amplifying the musical tone signal supplied from the DAC 8 and emitting it.

  The microphone 10 inputs instrument sounds from natural instruments (for example, guitars, pianos, etc.) emitted in response to actual manual performance operations (actual performances) by the performer. The waveform signal of the input instrument sounds is It is sent to the ADC 11. The ADC 11 is an analog-digital signal converter that converts the analog waveform signal of the supplied instrument sound into digital phrase waveform data. The converted phrase waveform data is recorded on the hard disk 13 under the control of the recording circuit 12. The The recorded phrase waveform data is given predetermined attribute information as will be described later, and each data analysis is performed based on the pattern performance data reproduced at the time of recording the actual performance by the sound source 6. Pattern analysis data (see FIG. 5 described later) corresponding to each sound (partial waveform) is given. Each process such as recording the phrase waveform data by recording the actual performance described above, and adding attribute information and pattern analysis data to the recorded phrase waveform data is performed by the CPU 1 as a “data generation process” program (FIG. 6). In addition, in order to be able to use the generated phrase waveform data, the phrase waveform generated by the performer with respect to the map data for arpeggio performance in which the phrase waveform data to be used is mapped for each arpeggio pattern is used. Data must be registered in advance. A detailed description thereof will be described later. The reproduction circuit 14 reads the phrase waveform data to be reproduced from the hard disk 13 in accordance with the designation of the arpeggio pattern or chord, and generates performance waveform data using this. The performance waveform data generated by the reproduction circuit 14 realizes an arpeggio performance and is emitted from the sound system 9 via the mixer 7 and the DAC 8.

  The MIDI interface 15 inputs performance data (MIDI data) in MIDI format from another externally connected MIDI device or the like (not shown) to the arpeggiator, or performance data (MIDI data) in MIDI format from the arpeggiator. Is an interface for outputting to other MIDI devices. Other MIDI devices may be any devices that generate MIDI data in response to user operations, and include any type of controls such as keyboard type, stringed instrument type, wind instrument type, percussion instrument type, and body-mounted type. The device may be an operation form). Note that the MIDI interface 15 is not limited to a dedicated MIDI interface, and the MIDI interface 15 is configured using a general-purpose interface such as RS-232C, USB (Universal Serial Bus), IEEE1394 (I-Triple 1394). You may make it do. In this case, data other than MIDI event data may be transmitted and received simultaneously. When a general-purpose interface as described above is used as the MIDI interface 15, other MIDI devices may be able to transmit and receive data other than MIDI event data. Of course, the data format of the performance data is not limited to data in the MIDI format, and may be in other formats. In this case, the MIDI interface 15 and other MIDI devices are configured accordingly. The other interface 16 is a communication interface connected to a wired or wireless communication network such as a LAN, the Internet, or a telephone line, and is connected to a server computer (not shown) via the communication network and controlled from the server computer. Load programs or various data into the arpeggiator. For example, when the control program and various data are not stored in the ROM 2 or the like, it is used for downloading the control program and various data from the server computer. Such other interfaces 16 may be both wired or wireless and may include both.

  Next, in the arpeggio performance apparatus (arpeggiator) according to the present invention, processing for recording phrase waveform data and processing for adding attribute information and pattern analysis data to the phrase waveform data will be described with reference to FIGS. To do. First, recording of phrase waveform data accompanying recording of a real performance will be described with reference to FIG. FIG. 2 is a conceptual diagram showing an aspect when phrase waveform data is recorded in the arpeggio performance device according to the present invention.

  When recording the phrase waveform data in the arpeggiator according to the present invention, as shown in FIG. 2, first, the automatic performance means 21 is read into the ROM 2 or RAM 3 according to a desired arpeggio pattern specified in accordance with the operation of a pattern selection switch or the like. An exemplary arpeggio performance is automatically played by supplying exemplary pattern performance data (see FIG. 3 to be described later) including one or more phrases generated from an exemplary pattern performance data generation unit 20 composed of This phrase includes notes having characteristic information that realizes a designated arpeggio pattern. The automatic performance means 21 is mainly composed of the CPU 1 and the sound source 6 shown in FIG. Then, the CPU 1 executes a known automatic performance program (not shown) to generate a sound source parameter based on the pattern performance data 20 read from the ROM 2 or the RAM 3 and supply the sound source parameter to the sound source 6. 6 generates performance waveform data. The performance waveform data generated by the sound source 6 is emitted from the headphones 26 connected to the sound system 9 via the mixer 7 and the DAC 8. Since the performer 22 is wearing the headphones 26, the performer 22 can listen to the performance sound based on the model pattern performance data.

  Here, the pattern performance data for realizing the above-described exemplary arpeggio performance will be briefly described with reference to FIG. FIG. 3 is a conceptual diagram for explaining an embodiment of the data structure of pattern data. The pattern data shown in FIG. 3 includes a pattern ID consisting of a unique number assigned to each predetermined arpeggio pattern, and an example of one or more phrases in a MIDI format such as SMF (Standard MIDI File). The pattern performance data, which is sequence data representing the arpeggio performance, is stored in the ROM 2, RAM 3, and the like. The pattern performance data is a combination of timing data and event data. The event data is data related to performance event information such as a note-on event for instructing the sound generation and a note-off event for instructing the sound mute. This event data is used in combination with timing data. The timing data in this embodiment is time data indicating a time interval from event data to the next event data (that is, duration data), but is not limited to this, and is relative time from a specific time or absolute time. Any format such as using the time itself may be used. In short, the pattern performance data format is the “event + absolute time” format in which the event occurrence time is expressed in absolute time within a song or measure, and the event occurrence time is expressed in terms of the time from the previous event. "Event + relative time" format, "pitch (rest) + note length" format that represents performance data in terms of note pitch and note length or rest and rest length, minimum performance resolution Any format may be used, such as a “solid method” format in which a memory area is secured for each time, and events are stored in a memory area corresponding to the time at which performance event information occurs. The pattern performance data may include various sound source control data (for example, data for controlling the volume) in addition to the event data and timing data.

  Returning to the description of FIG. 2, the performer 22 listens to the performance sound of the automatic performance (automatic accompaniment) based on the above-described exemplary pattern performance data, and actually uses a natural instrument such as a guitar following the performance sound. Play an arpeggio. Since the arpeggio performance by the performer 22 is performed while listening to the performance sound emitted from the headphones 26, the arpeggio performance performed following the performance sound is a natural sound having a similar arpeggio pattern. . A manual arpeggio performance waveform (referred to as a phrase waveform) is supplied to the waveform recording means 23 via the microphone 10, sampled into digital phrase waveform data, and composed of a hard disk 13 or the like for each phrase. Is recorded in the phrase waveform data recording unit 24. In this case, since the clocks of the automatic performance means 21 and the waveform recording means 23 are synchronized by the clock synchronization line 25, the phrase waveform data is recorded in synchronization with the performance timing. In this synchronization, the sampling clock of the waveform recording means 23 and the operation clock of the automatic performance means 21 are made common, and the timings of the automatic performance start and the recording start are made to coincide, or the difference between both timings is stored. ing. Thereby, the synchronization over the entire section of the phrase waveform data recorded in the phrase waveform data recording unit 24 and the pattern performance data to be automatically played is ensured. Therefore, even if the tempo of the pattern performance data is changed in the middle, the synchronization is not affected at all. Since the performance sound that is automatically reproduced during recording is generated based on the pattern performance data whose pitch is corrected in accordance with the designated “chord”, such automatic reproduction is referred to as “automatic accompaniment (reproduction)”. "

  The waveform recording means 23 and the phrase waveform data recording unit 24 are constituted by the CPU 1, ADC 11, recording circuit 12, and hard disk 13 shown in FIG. A phrase waveform, which is a musical instrument sound input from the microphone 10 under the control of the CPU 1, is sampled by the ADC 11 and converted into digital phrase waveform data. The phrase waveform data is written and recorded in a predetermined storage area of the hard disk 13 by the recording circuit 12. As an aspect of the automatic performance, it is preferable that the performance is a solo part performance composed of pattern performance data of only the arpeggio part (part where the arpeggio performance) is desired to be recorded, but is not limited thereto. For example, it may be an all-part performance composed of a plurality of pattern performance data including the arpeggio part to be recorded of the phrase, or minus one based on the pattern performance data of one to a plurality of parts excluding the arpeggio part to be recorded for the phrase. It may be a performance. The performer 22 can perform the arpeggio part while listening to any of the performance sounds, and record the arpeggio part phrase. Note that because of the phrase division described later, even in the case of a minus one performance, the pattern performance data corresponding to the minus one performance must include the performance data of the arpeggio part.

  Next, FIG. 4 schematically shows the relationship between the pattern performance data that is automatically played and the phrase waveform data that is recorded along with the actual performance. FIG. 4 is a schematic diagram showing the relationship between pattern performance data and phrase waveform data. However, as shown in the upper part of FIG. 4, the pattern performance data shown in this embodiment includes the chord constituent sounds “C (do)”, “E (mi)” and “G (so)” of the chord “C (major)”. A predetermined arpeggio pattern is realized, and here, it is represented by notes on a staff. Phrase waveform data recorded when a performer actually performs an arpeggio performance using a natural instrument in accordance with the performance sound while listening to the performance sound generated based on the pattern performance data, It is a series of waveforms in which the waveforms of a plurality of sounds as shown in the lower part of FIG. 4 are temporally continuous.

  According to the present invention, attribute information to be described later is added to the phrase waveform data recorded as described above, and by analyzing the recorded phrase waveform data, the sound generation timing, pitch, etc. Various feature data related to the arpeggio are extracted, pattern analysis data is generated based on the extracted feature data, and the pattern analysis data is added to the phrase waveform data to manage them as a set of phrase data. The arpeggiator according to the present invention executes an arpeggio performance according to this phrase data. Therefore, the data structure and data generation procedure of such phrase data will be described. First, the data structure of phrase data will be described with reference to FIG. FIG. 5 is a conceptual diagram showing an example of the data structure of phrase data.

  As shown in FIG. 5, the phrase data includes attribute information such as a phrase ID, a pattern ID, a chord route, and a chord type, pattern analysis data, and phrase waveform data. The “phrase ID” is a unique number that is appropriately added every time the phrase waveform data is recorded. For example, a unique serial number is appropriately assigned in the order of recording. “Pattern ID” indicates what kind of arpeggio pattern the phrase waveform data is played and recorded, and is the pattern data (see FIG. 2) used when recording the phrase waveform data. The same ID as the pattern ID is assigned. “Cord root” and “chord type” indicate the root sound and chord type of the chord used when recording the phrase waveform data. The attribute information may further include “tempo” which is a tempo at the time of recording. “Pattern analysis data” is data that is added based on the analysis of recorded phrase waveform data. As will be described later, pitch information and start / start information for each note (each partial waveform) obtained by analyzing phrase waveform data. This is characteristic information for each divided note, such as an end point (start address and end address of each partial waveform in the phrase waveform database). Since this “pattern analysis data” is characteristic information obtained by analyzing the recorded phrase waveform data, it does not necessarily match the pattern performance data (see FIG. 3) used when recording the phrase waveform data. Rather, it will be somewhat different data. For example, when the performer manually plays a part of the pitch of the arpeggio pattern by mistake or when the performance timing is shifted, the pattern analysis data is different from the pattern performance data. “Phrase waveform data” is recorded phrase waveform data. In the phrase data, the actual phrase waveform data itself is not stored, but management information that is path information indicating the data storage area where the phrase waveform data is actually stored is stored. The corresponding phrase waveform data may be read from the data storage area that actually stores the phrase waveform data.

  Next, the procedure for generating the phrase data (phrase data generation process) will be described collectively. FIG. 6 is a flowchart illustrating an example of phrase data generation processing. It is assumed that the tempo is designated in advance before executing this process. In step S1, an arpeggio pattern (pattern ID) and chord (root sound and chord type) to be recorded are designated using a keyboard or the like. In step S2, the arpeggio sound with the designated chord is automatically reproduced based on the pattern performance data of the designated arpeggio pattern and the designated chord at the designated tempo, and the sound source 6 (for example, Generated using a waveform memory sound source. That is, the sound of each event data included in the pattern performance data of the specified arpeggio pattern is used by using the chord constituent sound according to the specified chord so that it can be referred to as the standard (exemplary) arpeggio performance when recording the phrase waveform data. The height is corrected, and the arpeggio performance sound that serves as an example according to the corrected pattern performance data is played automatically (accompaniment). Thus, the performer can listen to this model arpeggio performance. Along with the model arpeggio performance played in this way (imitation), the performer manually performs the arpeggio performance on the instrument in real time. In step S3, a manual arpeggio performance is recorded, and this is recorded as phrase waveform data. In step S4, attribute information (see FIG. 5) is added to the recorded phrase waveform data, with the pattern, code, tempo, etc. specified in step S1 as the pattern ID, chord root and type, and tempo. The recorded phrase waveform data is analyzed, the result is added as pattern analysis data, and this is stored in the hard disk 13 as phrase data.

  Here, the data analysis (refer to said step S4) performed in order to add said "pattern analysis data" is demonstrated. The automatic performance processing means 21 shown in FIG. 2 performs automatic performance (automatic accompaniment), and the waveform recording means 23 performs sampling of instrument sounds by the actual performance of the performer 22 in accordance with the automatic performance. The recorded phrase waveform data always includes performance data (corresponding performance data) for waveform division corresponding to the arpeggio performance to be recorded. The corresponding performance data is the same as the performance data of the arpeggio part included in the pattern performance data that was automatically played (automatic accompaniment) when the phrase waveform data was recorded. Performance data with corrections. Therefore, the phrase waveform data is divided into partial waveforms corresponding to the notes of the corresponding performance data according to the corresponding performance data (and the analysis result of the phrase waveform data itself). Next, data (pattern analysis data) such as pitch, tone length, and intensity corresponding to each divided partial waveform is given by referring to the performance information of each note of the corresponding performance data. In this case, since the phrase waveform data is synchronized with the pattern performance data, it can be divided into partial waveforms only by the sounding timing of the corresponding performance data. Further, the phrase waveform data may be subjected to frequency analysis (formant analysis, FFT analysis, etc.) to correct data such as the division position, pitch, tone length, intensity, and the like. That is, the start position of the formant of the partial waveform is searched in the vicinity of the provisional division position provisionally determined according to the pattern performance data, and the detected start position can be determined as the division position. In this way, it is possible to divide at a musically accurate position and to assign data such as accurate pitch, pitch, intensity, etc., compared to dividing based only on the analysis result of phrase waveform data. become.

  An example of dividing the phrase waveform data in this way is shown in FIG. In this example, in order from the beginning, the partial waveform a is the eighth note “C (de)” of the corresponding performance data, the partial waveform b is the eighth note “E (mi)”, and the partial waveform c is the eighth note. “G (So)”, partial waveform “d” is an eighth note “E (Mi)”, partial waveform “e” is a quarter note “G (So)”, and partial waveform “f” is a quarter note “E (Mi)”. ”Respectively. The pitch information, start / end point (address), etc. of each partial waveform thus divided are added as pattern analysis data. As described above, since the pattern performance data and the phrase waveform data are synchronized, each partial waveform matches the performance timing of each note of the corresponding performance data, and the length of the sound corresponding to the note. This is the waveform. Furthermore, instead of recording the musical instrument sound of a natural instrument that was manually played as a single note, the instrument sound of the phrase from a natural instrument that was played manually was recorded while listening to the performance sound of the phrase based on the pattern performance data. Therefore, it goes without saying that the tone has a natural tone.

  More specifically, in the dividing method using only the corresponding performance data, the pattern performance data to be automatically played and the phrase waveform data are divided using the fact that they are completely synchronized. Therefore, according to the data of each note included in the corresponding performance data, the time range corresponding to each note (until the sound after the release start (note off) from the start timing (note on) of the sound corresponding to the note is attenuated, or Phrase waveform data in the time range until the next note begins) can be cut out as a partial waveform. However, since the phrase waveform data is a sample of the arpeggio performance actually performed by the performer himself, the start timing of the note of the corresponding performance data and the start timing of the waveform corresponding to the note of the phrase waveform data do not necessarily exist. In some cases, it may be shifted. Therefore, based on the start timing of each note of the corresponding performance data, the waveform of the phrase waveform data before and after that (for example, several seconds before and after the start timing) is analyzed, and the start timing of the note in the phrase waveform data is detected and divided position If the correction is made, the partial waveform can be accurately cut out.

  Formant analysis and FFT analysis can be adopted as a waveform analysis method used when correcting the division position. In this formant analysis, first, an LPC (Linear Prediction Coding) coefficient is calculated from a correlation function of waveform data before and after the phrase waveform data with reference to the start timing of each note of the corresponding performance data. Next, the LPC coefficients are converted into formant parameters to obtain formants in the phrase waveform data. Then, the start timing of the note is detected from the rising position of the formant corresponding to the note (including the immediately preceding attack noise). Thereby, partial waveform data can be cut out at a musically accurate position and divided into partial waveforms corresponding to each note. In the FFT analysis, first, the fast Fourier transform of the phrase waveform data is performed while moving the time window for the preceding and following sections of the phrase waveform data with reference to the start timing of each note of the corresponding performance data. Next, the trajectory of the fundamental tone and the multiple harmonics corresponding to the note is detected, and the start timing of the note (including the rise of the attack noise immediately before) is detected from the rising position in the detected trajectory of the fundamental tone and the multiple harmonics. This method can also cut out partial waveforms at musically accurate positions and divide them into partial waveforms corresponding to each note. Also, since the pitch and intensity of each partial waveform can be determined from the detected fundamental tone and the trajectory of multiple harmonics, the partial waveform is corrected by correcting the pitch and intensity data of each note included in the corresponding performance data. Can be granted. Here, although formant analysis and FFT analysis have been described, other analysis methods such as pitch analysis and envelope analysis may be used for analysis for correcting data such as division positions, pitches, and intensity. .

  After recording the phrase waveform data as described above, the performer must previously register the generated phrase waveform data for the arpeggio performance map data in which the phrase waveform data to be used is mapped for each arpeggio pattern. There must be. Such map data will be described with reference to FIG. FIG. 7 is a conceptual diagram showing an embodiment of the data structure of map data.

  The map data is data for mapping composed of a map ID, a pattern ID, range number information, and a phrase ID for each range data divided into numbers corresponding to the range number information. The “map ID” is a unique number that is appropriately assigned to each created map data. “Pattern ID” is a pattern ID assigned to each arpeggio pattern, and information related to a plurality of pattern IDs can be registered in one map. For example, the “pattern ID” designates which arpeggio pattern is applied when the arpeggiator according to the present invention is driven. The “number of ranges” information is information indicating how many different ranges of phrase data are applied corresponding to the one pattern ID. For example, if “3”, different phrase data is used corresponding to each of the three ranges as the arpeggio pattern corresponding to the one pattern ID, or if “10”, each of the ten ranges is associated with each of the ten ranges. Use different phrase data. The plurality of phrase data used here is phrase data having the same pattern ID as the one pattern ID as attribute information. “Phrase ID” for each range data specifies phrase data (see FIG. 5) to be used for each range defined by each range data. As a method of determining the range, for example, there is a method of dividing the range by the “chord root”. That is, the root sound of the chord is divided into a plurality of ranges for each predetermined range, and the phrase waveform data to be applied is changed according to each range. For example, one octave is divided into ranges corresponding to three sound ranges such as “C to E ♭”, “E to G”, and “A ♭ to B”, and phrase waveform data applied to each range is made different. In general, for example, when phrase waveform data recorded with a code “C (major)” is applied and the code “G (major)” is to be realized by an arpeggiator, “C” is pitch-shifted to “G”. When the pitch shift amount is large, for example, a great change occurs in the timbre and tone quality, which is not convenient. In order to prevent such inconvenience, the phrase waveform data to be applied is changed for each range (range) of the root sound here. Note that the method of dividing the range is not limited to the “chord root sound” described above. For example, “chord type” or “chord velocity” (average value of velocity of each performance operator when operating multiple performance operators corresponding to chord constituent sounds), “tempo at recording” The above may be used as a reference, or may be divided into a plurality of ranges based on a combination of these appropriately, so that different phrase waveform data can be designated for each range. Here, when forming map data for mapping a plurality of phrase waveform data, the range of the phrase waveform data of each phrase data is represented by the “code root”, “code type”, “tempo” of the phrase data. It may be determined with reference to attribute information such as “”. By switching the phrase waveform data according to the range of chord, velocity, tempo, etc., it is possible not only to prevent deterioration of the timbre and tone quality as described above, but also to construct phrases according to chord, velocity, tempo, etc. You can add timbre and facial expression to each partial waveform.

  Next, the outline of the reproduction process executed by the reproduction circuit 14 of the arpeggiator shown in FIG. 1 will be described with reference to FIG. FIG. 8 is a block diagram for explaining the outline of the waveform data reproduction processing executed by the arpeggio performance device (arpeggiator) according to the present invention. In FIG. 8, the arrows in the figure represent the flow of data.

  First, the performer specifies map data to be used in advance according to the operation of the map selection switch (not shown). Next, the performer operates the operator 5 such as a pattern selection switch to select a desired arpeggio pattern to be played, and also operates the operator 5 such as a chord designation switch corresponding to the keyboard or each chord. Enter the code. Then, performance information such as note-on, note-off, and other controller signals is output from the operator 5 to the chord detector 31 in accordance with the operation of the operator 5 such as the keyboard. The performance information from the keyboard is input to the chord detector 31 as MIDI information such as note-on, note-off, and control change. The chord detection unit 31 detects the chord root sound and chord type (further, the chord root sound range, chord velocity, etc.) based on the combination of the performance information, and this is detected as a phrase data selection unit 34 and an exemplary arpeggio pattern reproduction unit 35. Send to. On the other hand, operation information indicating which switch is operated from the operation element 5 is output to the operation detection unit 32 in accordance with an operation of the operation element 5 other than the above. In the operation detection unit 32, “pattern designation” information is sent to the phrase data selection unit 34 and the exemplary arpeggio pattern reproduction unit 35 according to the operation of the pattern selection switch, for example, according to the type of the operated operator 5. In response to the operation, “tempo instruction” information is sent to the tempo clock unit 37. The tempo clock unit 37 generates a tempo clock having the tempo indicated by the “tempo instruction” information. The phrase data selection unit 34 is based on the “code designation” information from the code detection unit 31, the “pattern designation” information from the operation detection unit 32, and the “tempo instruction” information, and the arpeggio phrase DB (database) in the hard disk 13. ) Phrase waveform data to be applied is acquired from 33. That is, a plurality of map data and phrase data are respectively stored in the arpeggio phrase DB 33, and from the “pattern ID” data set corresponding to the “pattern designation” information in the map data (see FIG. 7) designated in advance. Then, the chord root sound and chord type included in the “chord instruction” information, and the phrase ID of the range data including the tempo indicated by the “tempo instruction” information are determined. Then, the corresponding phrase data (see FIG. 5), that is, the phrase waveform data and the corresponding pattern analysis data are determined based on the determined phrase ID. The exemplary arpeggio pattern reproduction unit 35 is based on the “pattern designation” information from the code detection unit 31 and the “pattern designation” information from the operation detection unit 32, from the exemplary pattern data DB (database) 36 in the hard disk 13. Pattern performance data (see FIG. 3) having a corresponding “pattern ID” is acquired, and an arpeggio pattern equivalent to the standard arpeggio performance is generated. That is, pattern data is stored in the pattern data DB 36, pattern performance data of “pattern ID” corresponding to “pattern designation” information is acquired, and based on the root sound and chord type indicated by the “code designation” information. Then, the pitch of each event data included in the pattern performance data is corrected, and the corrected pattern performance data is output to the phrase reproduction unit 38 as a reference arpeggio pattern.

  The phrase playback unit 38 generates musical performance waveform data by performing musical tone synthesis using the phrase data selected by the phrase data selection unit 33 in accordance with the normative arpeggio pattern supplied from the model arpeggio pattern playback unit 35. Is output from the sound system 9 to output a musical tone of the arpeggio performance. At this time, instead of simply synthesizing the musical tone according to the phrase waveform data selected by the phrase data selection unit 33, pattern analysis data selected by the phrase selection unit 33, pattern performance data of the reference arpeggio pattern, and Based on the tempo clock from the tempo clock unit 37, musical tone synthesis is performed while correcting the phrase waveform data for each partial waveform (waveform for each sound constituting the phrase) to generate performance waveform data. Specifically, each partial waveform of the phrase waveform data is associated with an event of each note of the pattern performance data of the reference arpeggio pattern, and each partial waveform is assigned a pitch of the partial waveform indicated by the phrase analysis data. The pitch correction by pitch shift is performed so that the pitch of the corresponding note becomes the pitch, and the reproduction timing and tone length of the partial waveform are the same as the tone generation timing and tone length of the corresponding note in the performance with the tempo clock. The playback start / end timing is corrected so that That is, since the change state of the sound based on the selected phrase waveform data is in the pattern analysis data, a difference is taken for each sound in the normative arpeggio pattern and the sound (each partial waveform) of the phrase waveform data is The pitch and timing are corrected by the difference. In addition, since the intensity of each partial waveform is also recorded in the phrase analysis data, it is possible to further correct the intensity based on the velocity for each sound.

  For example, if the arpeggio performance that is the norm at the time of recording is a chord “C (major)”, the recorded phrase waveform data is also “C” “E” that is a chord component sound of the chord “C (major)”. “A” and “G” are used as arpeggio patterns, and as pattern analysis data, “C”, “E”, and “G” are recorded as the pitches of the partial waveforms in the order of performance of the constituent sounds. When the phrase waveform data formed by the chord constituting sound of the chord “C (major)” is used and the chord “Cm (minor)” is designated as the chord designation, the chord “Cm (minor)” Since the chord constituent sounds are “C”, “E ♭” and “G”, if the recorded phrase waveform data is used as it is, an arpeggio performance of a desired chord is not output. Therefore, a correction target and a pitch shift amount are obtained by taking a difference between each sound in the pattern analysis data and the model arpeggio pattern. In this example, only the sound “E” in the phrase waveform data is pitch shifted to “E ♭”. For example, when the phrase waveform data formed by the constituent sound of the above-mentioned chord “C (major)” is used and the chord designation “E (minor)” is designated, all of the conventionally known samplers are used. Are uniformly pitch-shifted by the same amount (in this case, two whole notes), so “C” is changed to “E” two full notes, “E” is changed to “A ♭” two full notes, “ The pitch is shifted from “G” to “B” two full-tones. However, the chord constituent sound of the chord “E (minor)” is “E”, “G”, “B”, and when the pitch is uniformly shifted as described above, the second sound “G” becomes “A ♭”. It becomes a major chord. Compared to this, in the arpeggiator according to the present invention, as in the conventional sampler, there is no change in using the phrase waveform data, but it was obtained individually for each sound (each partial waveform) of the phrase waveform data. Since correction is performed by shifting the pitch by the difference, “C” is changed to “E”, which is 2 full tones, “E” is changed to “G”, which is 1 semitone and 1 semitone, and “G” is 2 The pitch is shifted to “B” above, and an accurate arpeggio performance corresponding to the specified chord can be performed.

  Furthermore, the playback start / end timing is adjusted to the pattern data modeled in accordance with the tempo clock, and the playback start timing of each partial waveform using the time stretch technology, which is a well-known time axis expansion / compression technology. And adjust the playback length. In this case, the entire phrase waveform may be expanded / compressed with the time stretch technique using the time stretch technique, and the time stretch technique may be used with the time stretch technique to appropriately expand / compress only the specific note (partial waveform). You may do it. Since any known technique may be used as the time stretch technique, description thereof is omitted here. As described above, the arpeggiator according to the present invention reproduces each sound (each partial waveform) of the selected phrase waveform data while individually shifting the pitch according to the root sound and type of the input chord. To be able to. This makes it possible for a performer to easily perform more realistic and diverse arpeggio performances with a desired arpeggio pattern in which the chord changes according to the chord designation by preparing a small amount of phrase waveform data. .

When the pitch changes greatly due to the change of each partial waveform in the phrase waveform data, etc., the waveform data in a different transition state different from the phrase waveform data prepared in advance is appropriately inserted and played back. By doing so, the connection between the sounds is improved, and a natural musical tone may be output.
The chord designation may be performed from automatic performance data. That is, the arpeggiator can also be applied to an automatic performance device, and it is convenient that an arpeggio performance using a corresponding chord can be easily performed based on a chord designation from automatic performance data.

1 is a block diagram of a hardware configuration showing an embodiment of the overall configuration of an arpeggio performance device according to the present invention. It is a conceptual diagram which shows the aspect at the time of recording phrase waveform data in the arpeggio performance apparatus which concerns on this invention. It is a conceptual diagram for demonstrating one Example of the data structure of pattern data. It is the schematic which shows the relationship between pattern performance data and phrase waveform data. It is a conceptual diagram which shows one Example of the data structure of phrase data. It is a flowchart which shows one Example of a phrase data generation process. It is a conceptual diagram which shows one Example of the data structure of map data. It is a block diagram for demonstrating the outline | summary of the reproduction | regeneration processing performed with the arpeggio performance apparatus based on this invention.

Explanation of symbols

1. CPU, 1A, timer, 2. ROM, 3. RAM, 4. Display, 5. Control, 6. Sound source, 7. Mixer, 8. DAC, 9. Sound system, 10. Microphone, 11. ADC 12, recording circuit, 13 hard disk, 14 playback circuit, 15 ... interface, 16 other interface, 1D communication bus (data and address bus), 20 pattern performance data, 21 Automatic performance means, 22 · Performer, 23 · Waveform recording means, 24 · Phrase waveform data, 25 · Clock synchronization line, 26 · Headphone, 31 · Chord detection unit, 32 · Operation detection unit, 33 · Arpeggio phrase database, 34 -Phrase data selection part, 35-Model arpeggio pattern playback, 36-Pattern data database, 37-Tempo black Click, 38-phrase reproduction unit

Claims (4)

First storage means for storing a plurality of phrase waveforms each corresponding to the arpeggio pattern indicated by the pattern ID and divided into a plurality of partial waveforms;
Second storage means for storing characteristic information such as pitches and sound generation timings related to each of the divided partial waveforms of the phrase waveform, corresponding to each of the plurality of phrase waveforms;
Means for allowing a user to specify a pattern ID and a chord related to the arpeggio performance to be reproduced;
Means for generating a normative arpeggio pattern according to the designated pattern ID and code;
Each partial waveform of the phrase waveform corresponding to the specified pattern ID stored in the first storage unit is based on the feature information stored in the second storage unit and the generated reference arpeggio pattern. An arpeggio performance device comprising a reproducing means for correcting and generating performance waveform data of the arpeggio performance. further,
Means to let the user specify the tempo for the arpeggio performance to be played,
Means for generating a tempo clock of the specified tempo,
2. The arpeggio performance device according to claim 1, wherein the reproduction means corrects the reproduction timing and the sound length of each partial waveform based on the generated tempo clock. The first storage means stores a plurality of phrase waveforms corresponding to one pattern ID, and
Each phrase waveform stored in the first storage means is associated with a unique phrase ID,
The reproduction means is
A third storage means for storing map data indicating a phrase ID for each range related to the root sound or type of the chord corresponding to one pattern ID;
Means for obtaining a phrase ID corresponding to the designated pattern ID and corresponding to a range including the designated code, based on the map data of the third storage means;
2. The arpeggio performance apparatus according to claim 1, further comprising means for reading out a phrase waveform corresponding to the acquired phrase ID from the first storage means. The first storage means stores a plurality of phrase waveforms corresponding to one pattern ID, and
Each phrase waveform stored in the first storage means is associated with a unique phrase ID,
The reproduction means is
Third storage means for storing map data indicating phrase IDs for each range related to the code and the tempo in correspondence with one pattern ID;
Means for acquiring a phrase ID corresponding to the designated pattern ID and a range including the designated code and the designated tempo based on the map data of the third storage means;
The arpeggio performance device according to claim 2, further comprising means for reading a phrase waveform corresponding to the acquired phrase ID from the first storage means.

Images