JP6028844B2 - Musical sound synthesizer and program - Google Patents

Description

  The present invention relates to a musical tone synthesizing apparatus and program for synthesizing musical sound waveforms including a plurality of different pitches such as chords. In particular, the present invention relates to a technique for synthesizing a musical sound waveform that imitates the sound of a natural musical instrument having strings such as a piano and a guitar, taking into account the influence of coupled vibration of strings.

  Conventionally, waveform sample data (hereinafter simply referred to as waveform data) encoded by an arbitrary encoding method such as PCM (pulse code modulation), DPCM (differential PCM), or ADPCM (adaptive differential PCM) is previously stored in the waveform memory. 2. Description of the Related Art Musical tone synthesis apparatuses based on a so-called “waveform memory readout” method in which musical tone waveforms are synthesized by storing them and reading them in correspondence with a desired pitch are known. Further, in Patent Document 1 shown below, in a musical tone generator using a sine wave synthesis method, for example, when a plurality of different musical tones corresponding to a plurality of notes such as chords are simultaneously generated, each of the different pitches is described. It is described that a musical sound waveform is synthesized by modifying it so as to be close to a predetermined frequency based on the frequency of a harmonic that is close in the harmonic structure of the musical sound.

  By the way, a musical sound waveform for generating a chord in a conventional apparatus of “waveform memory reading” method is, for example, among a large number of waveform data sampled in accordance with an output corresponding to an operation for each key from a natural musical instrument piano. To simultaneously read out a plurality of waveform data corresponding to a plurality of chord constituent sounds keyed by the user, synthesize a single musical tone waveform for each chord constituent tone, and further, a single musical tone waveform synthesized for each chord constituent tone It is generated by adding.

  However, the chords sounded based on the musical sound waveform generated in this way can give the user the impression that the chord constituent sounds are not in harmony with each other and each of them singly and slammingly. However, it was far from a beautiful chord that could be emitted from a natural musical instrument piano that can give the user the impression that the chord components are in harmony with each other.

  As described above, when a musical sound waveform including a plurality of musical tones having different pitches such as chords is generated in a conventional device, musical tones that can give the user an impression different from the chords emitted from a natural musical instrument piano are generated. The The reason for this is that the above-described musical sound synthesizer does not take into account any influence on the musical sound caused by the “coupled vibration” that occurs when multiple keys are simultaneously played on a natural musical instrument piano. This is because a plurality of musical sound waveforms individually synthesized from the sampled waveform data are considered to be merely added.

  Non-Patent Document 1 shown below describes coupled vibration of a plurality of strings forming one key of a piano. According to this, a natural musical instrument piano emits a single musical tone by simultaneously vibrating a plurality of strings (2 to 3 strings, etc., depending on the musical note to be generated) according to a single keystroke, The multiple strings, each tuned to a frequency that is almost the same but not perfectly matched, interact with each other through the pieces, so that the frequencies generated by the vibration of each string come together. At the same time, a relatively loud part (referred to as a hitting sound or a hitting part in this specification) of a periodic vibration part (referred to as a decay part in this specification) of a string that starts immediately after the string is damped quickly. In other words, so-called “coupled vibration” is described.

  FIG. 20 shows an example of a musical sound waveform that reflects the characteristics of the musical sound that are seen when “coupled vibration” occurs. From FIG. 20, the musical sound waveform when “coupled vibration” is generated is compared with the musical sound waveform when “coupled vibration” is not generated (rough shape is indicated by a dotted line in the figure). The level (amplitude) decreases as a whole, and in particular, a sounding portion with a large volume level that continues for a while immediately after the start of waveform synthesis is a part of a musical sound waveform with a low volume level that follows the sounding portion (referred to as a reverberation portion). Can also be seen to decrease at a relatively larger rate.

  In view of the fact that a plurality of strings vibrate in a single key as described above, the inventor of the present application attenuates the sound wave-sounding portion of the musical sound waveform, and the harmonics of the strings associated with different keys are also coupled. As a result of vibration, multiple frequencies come together and the sounding part of the musical sound waveform decays quickly, so the natural instrument piano can conclude that it can produce very beautiful chords that resonate clearly. It came to derive.

Japanese Unexamined Patent Publication No. 07-319467

Wayne Reach, G. "2. Physics of piano strings ... Science of musical instruments. Nikkei Science, 1987, p.21-32

  However, in the conventional apparatus, for example, in the “waveform memory read-out” method, it is associated with different keys that can give the user the impression that the sounds of the natural instrument resonate cleanly and harmonize with each other. In addition, there is a problem that it is difficult to synthesize a musical sound waveform composed of a plurality of pitches by reproducing a coupled vibration caused by harmonics of one or more strings. That is, since the conventional device shown in the above-mentioned Patent Document 1 merely modifies only the frequency of harmonics, the harmonics of strings associated with different keys as described above are coupled vibrations. Although it is possible to simulate a plurality of frequencies associated with each other, it is because no consideration has been given to imitating that the hit sound part attenuates quickly. The conventional device described in Patent Document 1 described above is originally a sine wave synthesis method and has a more complex structure for musical tone synthesis than a “waveform memory read” method device, and the hitting sound attenuates quickly. It is very difficult to add a configuration for imitating, and even if such a configuration can be added, the apparatus becomes complicated and large-sized and more expensive, which is inconvenient.

  The present invention has been made in view of the above-mentioned points, and a sound wave-sounding portion in which a plurality of harmonic frequencies composing a musical sound generated in accordance with the operation of individual performance operators (for example, keys) are combined with each other. It is possible to synthesize a musical sound waveform that includes multiple musical tones with different pitches, imitating so-called “coupled vibration of strings” (hereinafter simply referred to as coupled vibration). An object is to provide a musical tone synthesizer and a program.

  The musical tone synthesizer according to the present invention includes a musical tone generator that simultaneously generates a plurality of different musical tone forms corresponding to notes, and a harmonic structure database that stores musical tone harmonic structures corresponding to each note. Based on the information indicating the harmonic structure of each of the plurality of different tone pitches, and on the basis of the information indicating the acquired harmonic structure, the frequency of the harmonics constituting each of the plurality of musical sound waveforms to be generated is Band specifying means for specifying one or more arbitrary bands including a plurality of adjacent harmonics, and one or more of the specified bands among the harmonics constituting each of the plurality of musical sound waveforms to be generated Level adjusting means for attenuating the volume level of each overtone.

  According to the present invention, a plurality of musical tone waveforms having a pitch corresponding to each note are generated simultaneously for each of a plurality of notes. From the overtone structure database storing the overtone structure of the musical sound waveform corresponding to each note, obtaining information indicating the overtone structure of each of the plurality of different sound waveforms, and based on the obtained information indicating the overtone structure, One to a plurality of bands including overtones whose frequencies are close to each other among the overtones constituting each of the plurality of generated musical sound waveforms are specified. Based on the specified one or more bands, among the harmonics constituting each of the plurality of musical sound waveforms to be generated, the harmonics included in each of the bands, that is, the harmonics of which frequencies are close to each other. Decrease level. That is, the sound volume level of only the harmonics belonging to a specific band (in other words, frequency components that can be coupled and vibrated close to each other) among the harmonic forms having different pitches corresponding to the note are set. Decrease. By doing this, when synthesizing a musical sound waveform that contains multiple musical notes of different pitches corresponding to multiple notes, the musical sound that reflects the characteristics of the musical sound that is seen when "coupled vibration" occurs It is possible to generate a characteristic musical sound waveform having a reduced volume level compared to a case where no waveform, that is, “coupled vibration” is generated.

  In addition, since it is configured to obtain information indicating the harmonic structure of each of the plurality of musical sound waveforms having different pitches from a harmonic structure database that stores the harmonic structure of the musical sound waveform corresponding to each note, a plurality of musical sounds to be generated The overtone structure of the waveform can be grasped easily and quickly. In addition, simply by controlling the volume level of a specific overtone, it is possible to synthesize a musical sound waveform that includes multiple musical tones corresponding to multiple notes reflecting “coupled vibration”. However, there is also an advantage that the apparatus is not complicated and large-sized and is not expensive.

  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.

According to the present invention, the volume level is adjusted with respect to only specific harmonics whose frequencies are close to each other among musical tone waveforms having different pitches corresponding to the notes. It becomes possible to synthesize musical sound waveforms that include multiple different musical pitches corresponding to multiple notes that mimic the characteristics of musical sounds that can be seen when there is a decrease in “coupled vibration”. The effect of is obtained.
In addition, when generating musical sounds of stringed instruments, especially piano sounds, it is possible to generate a musical sound waveform having musical characteristics such as an actual piano sound by imitating “coupled vibration”. Although this alone is effective, it is also possible to make chords resonate cleanly by imitating “coupled vibration” even for sounds other than stringed instruments and artificial sounds. New effects that could not be obtained.

It is a hardware block diagram which shows one Example of the whole structure of a musical tone synthesizer. It is a functional block diagram for demonstrating the outline | summary of 1st Example of a musical tone synthesis function. It is a conceptual diagram which shows an example of each data structure of a note list, a partial sound list, and a band list. It is a conceptual diagram which shows an example of a data structure of a harmonic structure database. It is a flowchart which shows one Example of a musical tone synthesis process. It is a flowchart which shows one Example of a note-on process. It is a flowchart which shows one Example of a note-off process. It is a waveform diagram showing an example of a tone waveform synthesized by adding an attack waveform, a decay waveform, and an attack waveform and a decay waveform. It is a functional block diagram which shows the 2nd modification of a musical tone synthesis function. It is a functional block diagram which shows the 3rd modification of a musical tone synthesis function. It is a functional block diagram for demonstrating the outline | summary of 2nd Example (4th modification) of a musical tone synthesis function. It is a functional block diagram which shows an example of a level correction part. It is a functional block diagram which shows the 5th modification of a musical tone synthesis function. It is a flowchart which shows a part of note-on process which implement | achieves the 4th and 5th modification of a musical tone synthesis function. It is a flowchart which shows a part of note-off process which implement | achieves the 4th and 5th modification of a musical tone synthesis function. It is a functional block diagram which shows the 6th modification of a musical tone synthesis function. It is a functional block diagram which shows the 7th modification of a musical tone synthesis function. It is a flowchart which shows a part of note-on process which implement | achieves the 6th and 7th modification of a musical tone synthesis function. It is a flowchart which shows a part of note-off process which implement | achieves the 6th and 7th modification of a musical tone synthesis function. It is a wave form diagram which shows an example of the musical tone waveform reflecting the characteristic on the musical tone seen when the coupled vibration has arisen.

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

  FIG. 1 is a block diagram showing a hardware configuration example of an electronic musical instrument to which a musical tone synthesis apparatus according to the present invention is applied. The electronic musical instrument shown here performs musical tone synthesis by the “read out waveform memory” method, and performance information (note-on event, note-off event, etc.) created as needed by the performer operating the performance operator 4. (Including various control data such as event data, dynamics information, pitch bend information, etc.) in the pre-created automatic performance data that is electronically generated or sequentially supplied with execution of automatic performance processing, etc. Has a musical tone synthesis function for automatically and electronically generating musical tone based on the performance information included.

  This musical tone synthesis function selects waveform data to be used from among a large number of waveform sample data (waveform data) stored in the waveform memory based on performance information, and generates musical sound waveforms according to the selected waveform data. In this embodiment, the musical sound waveform is synthesized in consideration of the influence of “coupled vibration”. Such a tone synthesis function will be described later. The electronic musical instrument shown in this embodiment may have hardware other than that shown here. Here, a case where the minimum necessary resources are used will be described.

  The electronic musical instrument shown in FIG. 1 is configured using a computer, in which the musical tone synthesis process for realizing the musical tone synthesis function (see FIG. 5 described later) is a predetermined program for the computer to realize each process. This is implemented by executing (software). Of course, this processing is not limited to the form of computer software, but can be implemented in the form of a microprogram processed by a DSP (digital signal processor), and is not limited to the form of this type of program. You may implement in the form of the dedicated hardware apparatus comprised including the integrated circuit or the large-scale integrated circuit.

  In the electronic musical instrument shown in this embodiment, various processes are executed under the control of a microcomputer comprising a microprocessor unit (CPU) 1, a read only memory (ROM) 2, and a random access memory (RAM) 3. It has become. The CPU 1 controls the operation of the entire electronic musical instrument. For this CPU 1, ROM 2, RAM 3, performance operator 4, display 5, setting operator 6, sound source 7, storage device 8, and communication interface 9 are provided via a communication bus 1D (for example, data and address bus). Each is connected.

  The ROM 2 stores various programs executed by the CPU 1 or waveform data corresponding to various instrument sounds such as waveform data sampled in accordance with the output of each key from a natural musical instrument piano as a so-called waveform memory. To do. 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 performance operator 4 is a keyboard or the like having a plurality of keys for selecting musical notes, and has a key switch corresponding to each key, for example, a key pressing speed or a release key. A velocity representing the key speed can be detected. The performance operator 4 can be used not only for manual performance of musical sounds by hand-playing by the performer itself, but also as input means for selecting automatic performance data prepared in advance. Of course, the performance operator 4 is not limited to the form of a keyboard or the like, and needless to say, it may be of any form such as a neck with a string for selecting musical notes. . The display 5 is a display composed of, for example, a liquid crystal display panel (LCD) or a CRT, and displays various screens as well as various performance parameters such as timbres and effects used during performance. Various information such as a setting state, a waveform data list or an automatic performance data list, and a control state of the CPU 1 can be displayed. The performer can easily set various performance parameters, select automatic performance data, and the like by referring to the various information displayed on the display 5.

  Setting operators (switches, etc.) 6 are various operators, such as a selection switch for selecting automatic performance data to be automatically played, and a setting switch for setting various performance parameters such as timbre and effect used during performance. It is comprised including. Of course, to select / set / control the pitch, tone, effect, etc., operate the numeric keypad for numeric data input, the keyboard for character data input, or the pointer to set the position of various screens displayed on the display 5. Various operators such as a mouse may be included.

  The tone generator 7 can simultaneously generate musical tone signals on a plurality of channels, and provides performance information and / or arbitrary automatic information generated in response to the operation of the performance operator 4 given by the user via the communication bus 1D. A musical sound signal is generated by synthesizing musical sound waveforms based on performance information generated in response to the reproduction of performance data. In the electronic musical instrument shown here, for example, as a waveform data, the entire musical tone waveform of a single tone is stored in the waveform memory, or the musical tone of the decay portion consisting of the musical tone waveform of the attack portion consisting of an aperiodic component and the periodic portion of the single tone musical tone waveform. When the waveform data corresponding to the performance information is read from the waveform memory, the read waveform data is given to the sound source 7 via the bus line. It is buffered as appropriate. The sound source 7 synthesizes a musical sound waveform by outputting the waveform data stored in the buffer in accordance with a predetermined output sampling frequency, and performs control processing for taking into account the coupled vibration described later. The musical tone waveform subjected to the control process is given to the sound system 7A to be sounded. The sound system 7A includes, for example, a DA converter, an amplifier, a speaker, and the like, amplifies the musical sound waveform converted into an analog waveform by the DA converter, and emits the sound through the speaker.

  The storage device 8 stores various data such as the above-described waveform data and automatic performance data, various control programs executed by the CPU 1, and the like. When the control program is not stored in the ROM 2, the control program is stored in the storage device 8 (for example, a hard disk), and the control program is stored in the ROM 2 by reading it into the RAM 3. The CPU 1 can be operated as described above. In this way, control programs can be easily added and upgraded. The storage device 8 is not limited to a hard disk (HD), but can be attached in various forms such as a flexible disk (FD), a compact disk (CD), a magneto-optical disk (MO), or a DVD (Digital Versatile Disk). A storage device using a recording medium may be used. Alternatively, a semiconductor memory or the like may be used.

  A communication interface (I / F) 9 is an interface for transmitting and receiving various information such as a control program and various data described above between the electronic musical instrument and an external device (not shown). The communication interface 9 may be, for example, a MIDI interface, a LAN, the Internet, a telephone line, or the like, and may include both wired or wireless ones.

  Also in the electronic musical instrument shown in FIG. 1, the performance information generated when the performer operates the performance operator 4 or the SMF (Standard MIDI File) format prepared in advance, as in a known device. It has a musical tone synthesis function for generating musical tone waveforms based on performance information included in automatic performance data and the like. However, when synthesizing a musical sound waveform including a plurality of musical tones having different pitches, a plurality of waveform data stored in the waveform memory are simply output according to a predetermined output sampling frequency and added together. In addition to synthesizing musical sound waveforms that include multiple musical tones with different pitches, multiple different pitches that reproduce coupled vibrations by applying signal processing to take account of coupled vibrations described below. A musical sound waveform comprising the musical sounds is synthesized. This will be described below.

(First embodiment)
FIG. 2 is a functional block diagram for explaining the outline of the first embodiment of the tone synthesis function. The arrows in the figure represent the data flow. The keyboard 4 (performance operator) supplies performance information such as a note-on event and a note-off event to the sound source unit A and the determination unit D in accordance with a user's performance operation. The determination unit D is one function realized by the CPU 1. The performance information such as the note-on event and the note-off event includes a note number (note number) indicating a pitch previously associated with a key operated by the user, velocity information indicating a key pressing speed or a key releasing speed, etc. It is included.

  Here, an example is given in which performance information is supplied in accordance with the operation of the keyboard 4 as a performance operator. However, the present invention is not limited to this. For example, the performance information is created in advance according to an automatic performance (sequencer) function that can be realized by the CPU 1. The performance information included in the automatic performance data may be sequentially supplied, or the automatic performance data and the performance information may be sequentially supplied from an external device (not shown) via the communication interface 9. May be. In addition, when the damper pedal, one of the performance controls, is depressed, the sound continues even after the note-off event occurs. After that, the processing equivalent to the occurrence of the note-off event is performed by stopping the depression of the damper pedal. The description of the processing by the damper pedal operation will be omitted.

  When the determination unit D receives a note-on event or a note-off event from the keyboard 4, the determination unit D updates the note list, partial sound list, and band list that are temporarily stored in the RAM 3. FIG. 3 shows data structures of the note list, partial sound list, and band list.

  FIG. 3A shows the data structure of the note list. The note list stores / manages notes of musical sounds to be synthesized in each tone generation channel A1 of the sound source unit A by note numbers and the like. In this note list, the note number included in the note-on event is added as a note of a musical tone that starts synthesis of a musical tone waveform in response to the reception of a note-on event, while in response to the reception of a note-off event The note number included in the note-off event is deleted as the note of the tone that stops the synthesis of the tone waveform. Here, a note list in the case of synthesizing three different pitches corresponding to note number “45” (note A2), note number “48” (note C3), note number “55” (note G3). An example of

  FIG. 3B shows the data structure of the partial sound list, where the partial sound list is the frequency (Hz) of the fundamental tone and harmonics constituting the musical sound waveform to be synthesized, and the sound pressure determined in advance for each fundamental tone and harmonic. Store / manage percentage (db). Each time a new musical note (note number) is added to the above note list, a harmonic structure database (see FIG. 4), which will be described later, is referred to based on the added note, and synthesis is newly started in the partial sound list. The frequency and the sound pressure ratio of each of the fundamental tone and the overtone constituting the musical sound waveform are newly added.

  On the other hand, whenever a musical note (note number) is deleted from the above note list, the harmonic structure database is referred to based on the deleted note and included in the musical sound waveform in which synthesis is stopped from this partial sound list. The frequency and the sound pressure ratio of each of the fundamental tone and the harmonic overtone are deleted. Here, note numbers “45” (note A2), note numbers “48” (note C3), note numbers “55” (note G3), fundamental tone and secondary to eighth order constituting each of three different musical sound waveforms. An example of a partial sound list in the case where the frequency (Hz) and the sound pressure ratio (db) of each overtone are stored / managed is shown as an example. If these fundamental and overtones are shown in the frequency domain, FIG. It becomes like (D).

  Here, FIG. 4 shows the data structure of the harmonic structure database referred to when the partial sound list is created. The harmonic structure database is stored in the waveform memory, for example, for each original musical sound waveform (corresponding to waveform data stored in the waveform memory) generated in response to an operation of one key (performance operator) of the piano. It is a database in which the sound pressure ratio is stored in advance together with the fundamental and overtone frequencies constituting the musical sound waveform.

  As shown in FIG. 4, the database uses a note number as a search key, and has a harmonic structure including a fundamental tone constituting the original musical sound waveform corresponding to the waveform data stored in the waveform memory uniquely determined by the note number. Has in attribute information. As a harmonic structure constituting the musical sound waveform, a volume ratio (db) and a frequency (Hz) are stored for each fundamental tone and each harmonic. The volume ratio is a predetermined time interval of the original musical sound waveform uniquely determined by the note number (for example, from the beginning to the end of the musical sound waveform, or from 100 msec to 1000 msec after the waveform position where the overtone is sufficiently grown) Is the ratio of the volume of the entire musical sound waveform in the section to the volume of each overtone, expressed in decibels. The frequency is an average value of the frequencies of the harmonics constituting the musical sound waveform included in the musical tone waveform included in the predetermined musical time waveform of the original musical sound waveform uniquely determined by the note number.

These volume ratios and frequencies may be stored based on the result of frequency analysis of the original musical sound waveform obtained from a natural musical instrument piano. However, if there is a low range, there is a range that includes 100 or more overtones as overtones, but it is not necessary to store all of these overtones. It may be limited to information. For example, in the example shown in FIG. 4, attribute information from the fundamental tone to the 8th harmonic is stored.
In the example shown in FIG. 4, the harmonic structure of the musical sound waveform is formed over the entire key range (88 keys) to which 88 different notes corresponding to the notes “A0 to C8” generated from a standard piano are assigned. Although the stored example is shown, the present invention is not limited to this, and only the musical sound waveform overtone structure in the key range of the middle range frequently played by the user may be stored. Moreover, it is not necessary to memorize the harmonic structure of a musical sound waveform in the high key range where almost no overtones can be heard.

  In the example shown in FIG. 4, the waveform data corresponding to the waveform memory is stored, and the harmonic structure of the original musical sound waveform with different semitones corresponding to each key is changed for all keys. The example which memorized about the musical sound waveform was shown. However, in some cases, waveform data with different semitones corresponding to all keys may not be stored in the waveform memory. In that case, the waveform memory stores waveform data corresponding to an arbitrary original musical sound waveform acquired by sampling only one per octave. Then, by changing the playback speed of this arbitrary waveform data from 0.75 times to 1.5 times the original, it is possible to generate a plurality of musical sound waveforms of different notes of one octave using one waveform data. It has become. Even when the sound source has such a mechanism, the harmonic structure database changes the waveform data playback speed in addition to the musical tone harmonic structure obtained when the waveform data is reproduced at the original playback speed. It is also preferable to store the harmonic structure of the musical sound waveform obtained when the sound is reproduced, that is, to store the harmonic structure for each semitone.

  Alternatively, while the waveform memory stores several single-tone waveform data each corresponding to a certain note at discrete pitches, the harmonic structure database has one for each one of the waveform data. You may make it memorize | store a harmonic structure. In this case, when synthesizing the musical sound waveform of the specified note, the waveform data based on the specified note is specified and the harmonic structure is specified with reference to the harmonic structure database. The fundamental tone and the frequency of each overtone may be corrected according to the difference between the predetermined pitch corresponding to the specified note and the pitch of the specified waveform data.

  In the example shown in FIG. 4, it is assumed that the original musical sound waveform is uniquely determined by the note number. However, depending on the implementation, the same key number is operated to generate the same note number, but a plurality of waveform data are generated depending on the key pressing speed. Some of them select a waveform data and synthesize a musical sound waveform. For example, with respect to the key of the note “A2”, waveform data obtained by sampling a musical sound waveform of f (forte) generated when the player strongly plays the key, and generated when the player plays the key weakly. The waveform data sampled from the p (piano) musical sound waveform is stored separately in the waveform memory, and the waveform data obtained by sampling the f (forte) musical sound waveform according to the key press speed or the p (piano) musical sound waveform Some of them select one of the waveform data sampled. In such a case, the harmonic structure database takes note numbers and key press speeds as composite keys, and the original f (forte) tone waveform and p (piano) specified by the composite key (note number and key press speed). It is good to memorize the harmonic structure of each of the musical sound waveforms.

  In the example shown in FIG. 4, the harmonic structure database is shown as a relational database in which the note number and the original musical sound waveform harmonic structure are associated with each other. However, a database having another structure such as an object-oriented database may be used. Further, instead of being configured as a database, the association between the note number and the overtone structure of the original musical sound waveform may be modeled and configured by a function expressed by a mathematical expression.

  Returning to the description of FIG. 3, FIG. 3C shows the data structure of the band list. This band list is a predetermined number of harmonics (also referred to as partial sounds) stored in the partial sound list. Stores / manages what is extracted from the partial sound list, and the maximum attenuation level of the volume level in each band, which is a band that can cause so-called “coupled vibrations” that are close to each other in the band condition and can affect each other To do. In the present embodiment, on the condition that a plurality of overtones are close to each other with a predetermined frequency difference or less and a volume level of the plurality of overtones is a predetermined value or more, a frequency range formed by these overtones is a so-called “ream”. The band is extracted as a band that can generate “synthetic vibration” and stored in the band list. For example, a frequency of a certain harmonic is close to an adjacent harmonic frequency within 1.0%, and at least one volume level of the harmonics is, for example, “−70 dB” (predetermined threshold value) or more. In this case, the frequency range formed by these multiple overtones is extracted as a band that can cause “coupled vibration”. Here, when each key associated with the note “A2 (45)”, the note “C3 (48)”, and the note “G3 (55)” (note number in parentheses) is pressed, that is, FIG. A specific description will be given by taking as an example the case where the note list shown in FIG. 3 (A) and the partial sound list shown in FIG. 3 (B) are generated.

  First, consider the proximity of harmonic frequencies. For example, as shown in FIG. 3B, the seventh harmonic of “A2” and the fourth harmonic of “G3” are close to each other with a frequency difference within 1%, and the fourth harmonic of “G3”. And the sixth harmonic of “C3” are close to each other with a frequency difference of 1% or less. Therefore, in the figure composed of the seventh harmonic of “A2”, the fourth harmonic of “G3”, and the sixth harmonic of “C3”, the band indicated by band 3 (indicated by a circled number in the figure, the same applies hereinafter) is shown. Candidate band for generating “coupled vibration”. Similarly, the band indicated by band 1 in the figure composed of the second harmonic of “G3” and the third harmonic of “C3” is also a candidate, and the fifth harmonic of “C3” and the sixth harmonic of “A2”. A band indicated by band 2 in the figure composed of overtones is also a candidate.

  Next, the volume level of each overtone included in the candidate band is considered. Here, if the volume level of at least one overtone is “−70 dB” or more, a band including the overtone is extracted as a band that can cause “coupled vibration”. As described above, as a result of considering the proximity of harmonic frequencies, the bands indicated by bands 1 to 3 in FIG. However, with regard to band 1, the volume levels of both the second harmonic of “G3” and the third harmonic of “C3” are both below the threshold value of “−70 dB”. On the other hand, regarding the band 2 and the band 3, at least one volume level of a plurality of overtones included in these bands is equal to or higher than a threshold value of “−70 dB”. Therefore, since band 2 and band 3 satisfy the above conditions, they are extracted as bands that can generate “coupled vibration”, while band 1 does not satisfy the above conditions and is not extracted as a band that can generate “coupled vibrations”. .

In this way, since bands that can cause “coupled vibration” are extracted, the band list stores the frequencies from the minimum frequency to the maximum frequency among the frequencies of the multiple harmonics included in the extracted band. In the above example, as shown in FIG. 3C, the band of “1308.2 Hz to 1320.2 Hz” and the band of “1540.4 Hz to 1569.9 Hz” are stored in the band list as bands that can cause “coupled vibration”. .
It should be noted that the above-described method for extracting a band that can generate “coupled vibration” is an example, and another method may be used. For example, among the bands 1 to 3 indicated as candidates, a band in which the total value obtained by adding the volume levels of a plurality of overtones included in each band is larger than a predetermined threshold value is set as “coupled vibration”. It may be extracted as a possible band.

On the other hand, the maximum attenuation of the band list is determined by the number of bands that can cause the extracted “coupled vibration” and its bandwidth. If the number of overtones included in the extracted band is m (pieces), the bandwidth is b (Hz), and the frequency difference of 1%, which is one condition for generating coupled vibration, is d (Hz), the maximum Attenuation amount L (dB) is calculated according to Equation 1 shown below.
L = -3 ・ (m-1) / ceil (b / d) ・ ・ ・ Equation 1

  Explaining the meaning of Equation 1 above, it is assumed that m harmonics are scattered within a frequency difference d (Hz) between adjacent harmonics within a bandwidth b (Hz). Each of these harmonics is attracted to “ceil (b / d)” points within the bandwidth b (Hz) by the coupled vibration. If two overtones adjacent to a single point are attracted, it is considered that vibration energy is exchanged in a plurality of strings that emit the overtones, and a sound wave-shaped sounding portion (with a volume level generated soon after pronunciation) Large periodic components) are attenuated to a maximum of "-3dB" according to the signal level. Also, when 3 harmonics are attracted to a single point, it is attenuated to a maximum of "-6dB", and when 4 harmonics are attracted to a single point, it is attenuated to a maximum of "-9dB". The amount of attenuation L is determined.

  The maximum attenuation L is determined by the above equation 1 when a large number of harmonics are attracted to a certain frequency by a coupled vibration for one key of a natural musical instrument piano. The energy exchange should be done, and therefore the energy of the sound waveform sounding part is distributed to the reverberation part of the sound waveform emitted from more strings, so the sounding part attenuates faster. Based on the assumption that there will be. For example, for the band 3 (see FIG. 3B) extracted when the keys to which the notes “A2”, “C3”, and “G3” are respectively assigned are simultaneously pressed, “1540.4 Hz˜ Three harmonics in the “1569.9 Hz” band (seventh harmonic of A2, sixth harmonic of C3, and fourth harmonic of G3) are attracted to one point, so that −6 dB is determined as the maximum attenuation value L. In this way, the lists shown in FIGS. 3A to 3C are updated in accordance with the key operation.

  Note that when a new harmonic is added to the partial list in response to a new note (note number) being added to the note list upon receipt of a note-on event, a new note is added in the band where the harmonics are close May be wider than before. In such a case, a plurality of bands extracted as bands that may cause “coupled vibration” in the band list may be reconfigured to be combined into one wide band. On the other hand, when the corresponding overtone is deleted from the partial list in response to the corresponding note being deleted from the note list by receiving a note-off event, the band where the overtones are close becomes narrower than before the note is deleted. Sometimes. In such a case, one band that has been extracted as a band that can cause “coupled vibration” in the band list may be reconfigured so as to be divided into a plurality of narrow bands.

  Returning to the description of FIG. 2, in this embodiment, when a plurality of pieces of performance information are simultaneously supplied from the keyboard 4 to the sound source 7 in accordance with the operation of a plurality of keys, the sound source 7 is provided with different notes included in these pieces of performance information. For each number, waveform data uniquely determined by each note number is read out from a waveform memory (not shown), and finally one musical tone waveform is synthesized using the read out waveform data and output as an output signal. The sound source 7 can be roughly divided into a sound source unit A, a filter unit B, and a control unit C as shown in FIG. The tone generator unit A includes a plurality of tone generation channel units A1, and the tone generation channel unit A1 to be used from among the empty channels (that is, the tone generation channel unit A1 that is open) in response to the supply of note-on event data from the keyboard 4. , And a single tone waveform corresponding to the note is synthesized using the waveform data read from the waveform memory for each of the reserved tone generation channel portions A1. When a plurality of keys are operated by the user at the same time, a sound generation channel A1 is secured for each of a plurality of note event data (specifically note numbers) supplied from the keyboard 4, so that the plurality of sound generation channels secured. In the part A1, single musical tone waveforms corresponding to different notes (for example, A2, C3, G3) are respectively synthesized. Since a method for synthesizing a musical tone of a single tone using such waveform data is known, a detailed description thereof is omitted here.

  A single musical tone waveform synthesized by each sound generation channel unit A1 is added by the adding unit A2, thereby generating one musical tone waveform including a plurality of musical tones corresponding to a plurality of notes. One generated musical sound waveform is output to the filter unit B and the control unit C. When note-off event data (that is, the same note number) corresponding to the note-on event data is supplied from the keyboard 4, the tone generator A synthesizes a single tone waveform corresponding to the note number. The tone synthesis of the tone generation channel portion A1 is finished and the tone generation channel portion A1 is released.

  The filter unit B includes a plurality of, for example, Notch filters B1 connected in series. Each Notch filter B1 has a volume level in an arbitrary frequency band of a musical sound waveform output from the sound source unit A and an arbitrary attenuation ( It decreases by an attenuation amount L ′) described later. The attenuation amount of the frequency band and volume level controlled by each Notch filter B1 is instructed (output) from the control unit C and the determination unit D.

  For example, when a musical sound waveform including a plurality of musical tones corresponding to the note “A2”, the note “C3”, and the note “G3” is output from the sound source unit A, the determination is made. As shown in FIG. 3C, the part D extracts the band 2 and the band 3 as bands that can cause “coupled vibration”. Then, the determination unit D allocates the extracted band 3 to the second Notch filter B1 so as to allocate the extracted band 2 to the first Notch filter B1 close to the input side of the musical sound waveform of the filter unit B. While instructing each, it is instructed not to use each of the 3rd to Nth Notch filters B1, that is, whether all the remaining Notch filters B1 are set to flat characteristics or set to through. For example, in the second Notch filter B1, a Q value corresponding to the bandwidth of band 3, that is, “1540.4 to 1569.9 Hz” is instructed, and “1555.2 Hz” is instructed as the center frequency (F value) of the band 3. The

  Further, the determination unit D assigns the extracted band to the Notch filter B1 every time a band that may cause “coupled vibration” is extracted. At this time, the control unit C corresponds to each Notch filter B1 of the filter unit B. Among the plurality of sequence attenuation amount determination units, which will be described later, is provided with a sequence attenuation amount determination unit corresponding to the Notch filter B1 to which the band is allocated, and the band is determined with respect to the secured attenuation amount determination unit. A bandwidth (Q value) and a center frequency (F value) are expressed. The determination unit D also instructs the maximum attenuation amount L of the volume level to the secured attenuation amount determination unit.

  The control unit C instructs the attenuation L ′ of the sound volume level for each Notch filter B1 of the filter unit B to which the band is assigned by the determination unit D. The volume level attenuation L ′ includes the bandwidth (Q value) and center frequency (F value) indicated by the determination unit D for each of a plurality of series attenuation determination units, the maximum volume level attenuation L, and the sound source. It is determined in accordance with the volume level of the musical sound waveform output from the part A. As shown in FIG. 2, the control unit C corresponds to each Notch filter B1 of the filter unit B with an attenuation amount determination unit including a bandpass filter (BPF) C1, an envelope follower (EF) C2, and a multiplier (Conv) C3. And have multiple series. Note that the bandpass filter C1 and the Notch filter B1 only need to be configured by, for example, an IIR filter.

  In the attenuation determination unit of each series, the bandpass filter C1 is a band instructed by the determination unit D only to the tone waveform included in the band assigned to the corresponding Notch filter B1 among the tone waveforms output from the sound source unit A. Take out according to width (Q value) and center frequency (F value). The envelope follower C2 obtains an instantaneous volume level at that moment by performing envelope detection on the extracted musical tone waveform of the predetermined band. The multiplier C3 calculates a volume level attenuation L ′ by multiplying the obtained instantaneous volume level by the maximum attenuation L of the volume level instructed by the determination unit D. The volume level attenuation amount L ′ thus calculated fluctuates up to the maximum attenuation level L according to the volume level of the musical tone waveform at that time included in the band assigned to the Notch filter B1. Of course, the calculation method of the volume level attenuation amount L ′ by multiplication here is only an example, and the degree of closeness of overtones or the volume level of the input musical sound waveform can be reflected in the attenuation amount. It may be an arithmetic method. In this way, the control unit C obtains the attenuation amount L ′ based on the volume level of the musical sound waveform output from the sound source unit A, and controls the filter unit B based on the obtained attenuation amount L ′. The volume level of any one of the harmonics constituting the musical sound waveform output from the sound source unit A for each band that can generate “coupled vibration” can be appropriately reduced.

  Next, a control program for realizing the musical tone synthesis function in the electronic musical instrument described above will be described with reference to FIGS. FIG. 5 is a flowchart showing an embodiment of a tone synthesis process for realizing a tone synthesis function in an electronic musical instrument. This “musical tone synthesis process” is a process controlled by the CPU 1 and is started in accordance with the power-on operation of the electronic musical instrument by the user and repeated until the power-off operation is performed.

  Step S1 performs an initial setting process. In this initial setting, variables temporarily stored in the RAM 3 are set to an initial state, or a set value for each notch filter B1 of the note list, partial sound list, band list, filter unit B, and attenuation of each series of the control unit C Processing such as clearing a setting value for the amount determination unit is performed. Accordingly, all Notch filters B1 and bandpass filters C1 are initialized to flat characteristics or waveform through states.

  In step S2, based on the detection (reception) of note-on event data or note-off event data, either “note-on process” (step S3) or “note-off process” (step S4) is executed. Execute processing distribution according to event data. Note-on event data or note-off event data is detected as follows. Note-on event data is emitted when the user presses the keyboard 4 (performance operator), and note-off event data is issued when the user performs a key release operation. Detected event data. Alternatively, note-on event data or note-off event data emitted at any time according to the progress of musical tone by playing automatic performance data (not shown) that is processed as a separate task by the CPU 1 or the like is detected. . Furthermore, according to the reception of note-on event data and note-off event data transmitted from an external device connected via the communication interface 9, those event data are detected.

  In step S2, if note-on event data is detected, "note-on processing" (step S3) is executed, and if note-off event data is detected, "note-off processing" (step S4) is executed. After completion of these “note-on process” or “note-off process”, the process returns to step S2. Thus, the “note-on process” is executed every time the note-on event data is detected, and the “note-off process” is executed every time the note-off event data is detected.

  FIG. 6 is a flowchart showing an example of “note-on processing” (see step S3 in FIG. 5). In step S11, a note (specifically a note number) designated by the note-on event data is written in the note list (see FIG. 3A). In step S12, the overtone structure database (see FIG. 4) is searched using the written note as a search key. In step S13, the overtone structure obtained as a result of searching the overtone structure database is written in the partial sound list (see FIG. 3B). In step S14, a band that can generate “coupled vibration” that satisfies a predetermined condition is extracted based on the overtone structure written in the partial sound list. At this time, the maximum attenuation L is determined according to the above equation 1. In step S15, the band list (see FIG. 3C) is updated with the band that can cause the extracted “coupled vibration” and the determined maximum attenuation L.

  In step S16, control is performed to update the settings of the filter unit B and the control unit C (see FIG. 2) of the sound source 7 in accordance with the update of the band list. Specifically, a frequency band that can cause the extracted “coupled vibration” is set for the Notch filter B1 of the filter unit B together with its center frequency (F value) and bandwidth (Q value). Further, the bandwidth (Q value) and the center frequency (F value) are set for the attenuation determining unit of each series of the control unit C. At this time, the maximum attenuation L of the volume level is also instructed. In step S17, the sound source 7 (more specifically, the sound source part A) is reserved for the sound generation channel part A1 that is not in use, and the sound corresponding to the note (note number) stored in the note list. Instructs the start of synthesis of a high tone waveform. In addition, regarding the processing in steps S11 to S16, when the musical sound waveform is configured to be divided into an attack waveform and a decay waveform as in, for example, a second modification described later, these processes are performed only on the decay waveform. It is not necessary to perform the attack waveform.

  When the start of synthesis of musical sound waveforms is instructed, the sound source unit A synthesizes a single musical tone waveform for each tone generation channel unit A1, and adds the synthesized single musical tone waveforms to differ in correspondence with a plurality of notes. A musical sound waveform including the pitch is output. When a musical sound waveform including different pitches corresponding to a plurality of notes is output from the sound source unit A, the extracted “ream” is extracted by the Notch filter B1 of the filter unit B out of the harmonics constituting the musical sound waveform. With respect to overtones included in a frequency band that can generate “synthetic vibration”, control is performed to reduce the volume level according to the volume level of the musical sound waveform. At this time, the volume level attenuation L ′ fluctuates in accordance with the volume level of the musical sound waveform at that time, but the volume level attenuation L ′ increases relatively as the volume level of the musical sound waveform increases. As a result, the volume level of the sound waveform sounding portion can be greatly reduced as compared with the reverberation portion.

  FIG. 7 is a flowchart showing an example of “note-off processing” (see step S4 in FIG. 5). In step S21, the musical tone in the tone generation channel unit A1 that is synthesizing the musical tone waveform having the pitch corresponding to the note (note number) specified by the note-off event data for the sound source 7 (more specifically, the sound source unit A). A release instruction is issued to end the synthesis. In response to this release instruction, the sound source unit A performs a control to mute the sound level of the musical sound being sounded by gradually lowering it over a predetermined time. In step S22, it is determined whether or not the tone synthesis in the corresponding tone generation channel section A1 has been completed. If it is determined that the musical tone synthesis has not ended (NO in step S22), the process waits until the musical tone being sounded is completely muted (the process in step S22 is repeated). If it is determined that the tone synthesis has been completed (YES in step S22), the note (note number) that has been turned off is deleted from the note list (see FIG. 3A) (step S23).

  In step S24, the overtone structure corresponding to the deleted note is deleted from the partial sound list (see FIG. 3B). When a harmonic structure corresponding to an arbitrary note is deleted from the partial sound list, a band that may cause “coupled vibration” in the partial sound list may be reconstructed as described above. A step S25 extracts a band that can cause “coupled vibration” in the partial sound list. In step S26, the bandwidth list (see FIG. 3C) is updated with the extracted bandwidth. In step S27, the settings of the filter unit B and the control unit C (see FIG. 2) are updated according to the update of the band list. In this case, among the harmonics constituting the musical sound waveform including different pitches corresponding to a plurality of other notes excluding the note that has been note-off from the sound source unit A, the Notch filter B1 of the filter unit B uses the Notch filter B1. Control is performed to reduce the volume level of overtones included in a frequency band that can cause “coupled vibration” based on the updated band list.

  As described above, the musical tone synthesizer according to the present embodiment simultaneously generates a plurality of musical tone waveforms corresponding to each note for each of a plurality of notes, and each of the musical sound waveforms having a plurality of different pitches corresponding to the notes. Then, an arbitrary band including harmonics whose frequencies are close to each other is specified for a plurality of harmonics constituting each musical sound waveform. This arbitrary band includes one or more bands depending on the overtone structure of each musical sound waveform. Based on the specified one or more bands, among the overtones constituting each of the generated plurality of musical sound waveforms, each overtone included in each of the bands, that is, each overtone whose frequency is close to each other, its volume Adjust the level. That is, the sound volume level of only the harmonics belonging to a specific band (in other words, frequency components that can be coupled and vibrated close to each other) among the harmonic forms having different pitches corresponding to the note are set. adjust. By doing this, when synthesizing a musical sound waveform that contains multiple musical notes of different pitches corresponding to multiple notes, the musical sound that reflects the characteristics of the musical sound that is seen when "coupled vibration" occurs It is possible to generate a characteristic musical sound waveform having a reduced volume level compared to a case where no waveform, that is, “coupled vibration” is generated.

(First modification)
As mentioned above, although an example of embodiment was demonstrated based on drawing, this invention is not limited to this, It cannot be overemphasized that embodiment of various modifications is possible. For example, as a first modification, the determination unit is independent of the attenuation L ′ obtained according to the volume level of a musical sound waveform including a plurality of musical tones output from the sound source unit A as described above. The volume level of a plurality of overtones included in a specific frequency band in which the overtones are close may be simply decreased at the same rate by the maximum attenuation amount L calculated according to the above equation 1 using D.

  In this case, the control unit C is not included in the functional block diagram shown in FIG. 2, and the determination unit D has a bandwidth (Q value) representing a specific frequency band for each Notch filter B1 of the filter unit B and In addition to setting the center frequency (F value), the maximum attenuation L is also set directly. However, if the volume level of a plurality of overtones included in a specific frequency band is lowered at the same rate, the volume level of the reverberation part as well as the sound waveform sounding part is lowered. However, although the first modified example has a simple configuration that does not have the control unit C as compared with the above-described embodiment, as a result, it is possible to generate a musical sound in which the sound of a chord can be heard to some extent. It is advantageous because it can be done.

(Second modification)
In general, a musical sound waveform is synthesized first in terms of “musical waveform of an attack part composed of non-periodic components” (hereinafter referred to as an attack waveform) and “decay composed of periodic components” synthesized later in time. It is possible to divide it into a musical tone waveform (hereinafter referred to as a decay waveform).

  Here, FIG. 8 shows, in order from the top, an example of a musical sound waveform synthesized by adding an attack waveform, a decay waveform, and an attack waveform and a decay waveform when “coupled vibration” occurs. According to FIG. 8, when “coupled vibration” occurs, the attack waveform does not change (does not lower) the volume level when “coupled vibration” does not occur. It can be understood that the volume level of the hitting portion is lower than when no “vibration” occurs. This is because the “coupled vibration” is caused by the interaction of periodic vibrations between multiple strings, while the decay waveform is susceptible to “coupled vibration”, while the attack waveform is “coupled”. It can be said that this is a result that can occur because it is hardly affected by “vibration”.

  In view of the above, the second modification shown in FIG. 9 synthesizes the decay waveform and the attack waveform corresponding to each note separately for each note in the sound source unit (Aa, Ab), and finally, a plurality of notes. By adding the decay waveform and the attack waveform, one tone waveform including a plurality of pitches corresponding to a plurality of notes is generated, and the separately synthesized attack waveform and decay waveform Before adding, the effect of “coupled vibration” can be reflected only on the decay waveform. FIG. 9 is a functional block diagram showing a second modification of the musical tone synthesis function. However, the same components as those in the functional block diagram of the first embodiment shown in FIG. 2 are denoted by the same reference numerals and their operations are the same as described above, and detailed description thereof is omitted here.

  The second modified example shown in FIG. 9 is divided into a first sound source unit Aa in which the sound source unit synthesizes a decay waveform and a second sound source unit Ab in which an attack waveform is synthesized. Each of the first sound source unit Aa and the second sound source unit Ab is provided with a plurality of sound generation channel portions (Aa1, Ab1), and the sound generation channel used from the empty channels in response to the supply of the note-on event from the keyboard 4 Section (Aa1, Ab1) is secured, and the waveform data of the decay section or the waveform data of the attack section stored in the waveform memory is read for each secured sound generation channel section (Aa1, Ab1), and this is used to correspond to the notes The decay waveform or attack waveform is synthesized.

  Decay waveforms or attack waveforms synthesized by the sound generation channel portions Aa1 and Ab1 of the sound source portions Aa and Ab are added by the adding portions Aa2 and Ab2, respectively, thereby including a plurality of pitches corresponding to a plurality of notes. The decay waveform or attack waveform is finally sent to the output adding unit K and added, thereby including a plurality of pitches corresponding to a plurality of notes. A musical sound waveform is synthesized. However, the attack waveform output from the second sound source unit Ab is sent to the output adding unit K as it is, but the decay waveform output from the first sound source unit Aa is sent to the output adding unit K after passing through the filter unit B. .

  In the case of this embodiment, each Notch filter B1 of the filter unit B has a bandwidth (Q value), a center frequency (F value), and a volume level attenuation L indicated (output) from the control unit C and the determination unit D. According to ', the volume level of overtones in an arbitrary frequency band is reduced by an arbitrary attenuation amount only with respect to the decay waveform output from the first sound source unit Aa. At this time, the control unit C calculates the attenuation L ′ of the volume level according to only the volume level of the decay waveform after addition output from the first sound source unit Aa. In this way, the influence of the coupled vibration is reflected on the decay waveform output from the first sound source section Aa, while the influence of the coupled vibration is not reflected on the attack waveform output from the second sound source section Ab. I am doing so.

(Third Modification)
In addition, in view of the fact that “coupled vibration” is caused by the interaction of periodic vibrations between a plurality of strings, when another musical sound waveform corresponding to a certain note is already being synthesized earlier in time, yet another When the musical sound waveform corresponding to the note is started to be synthesized later in time, the influence of the coupled vibration that may occur when synthesizing the attack portion of the musical sound waveform that has been synthesized later in time is earlier in time. It is conceivable to reach the decay portion of the musical sound waveform that has been synthesized.

  Therefore, the third modification shown in FIG. 10 synthesizes the decay waveform and the attack waveform corresponding to each note separately for each note in the sound source sections Aa and Ab, and finally the musical tone having a pitch corresponding to a plurality of notes. By adding the decay waveform and the attack waveform that contain the sound, one musical sound waveform that contains multiple pitches corresponding to multiple notes is generated. This is similar to the second modification described above in that the effect of “synthetic vibration” is reflected. However, in this third modification, the volume level of overtones in an arbitrary frequency band by the filter unit B is arbitrarily set on the assumption that the notes of the decay waveform and attack waveform synthesized in the sound source units Aa and Ab are different. When reducing the amount of attenuation, not only the volume level of the preceding waveform, that is, the decay waveform whose synthesis start timing precedes in time, but also the subsequent tone waveform, that is, the synthesis start timing, which is later in time. This is different from the second modification described above in that the amount of attenuation is obtained in consideration of the volume level of the attack waveform of the note. FIG. 10 is a functional block diagram showing a third modification of the musical tone synthesis function. However, the same number is attached | subjected about the same structure as the functional block diagram shown in FIG. 9 here.

  The third modification shown in FIG. 10 is different from the second modification shown in FIG. 9 in that it includes an input addition unit G. The input adding unit G includes a decay waveform including a plurality of pitches corresponding to a plurality of notes output from the first sound source unit Aa and a plurality of notes corresponding to the plurality of notes output from the second sound source unit Ab. By adding the attack waveforms including the musical tones of the pitches, one musical sound waveform including the musical tones having a plurality of pitches corresponding to the plurality of notes is synthesized and input to the control unit C. That is, as in the embodiment shown in FIG. 2, a single tone waveform including a plurality of musical tones corresponding to a plurality of notes is input to the control unit C, and the control unit C Calculates the attenuation L ′ of the sound volume level for each specific band based on the sound volume level of this entire musical sound waveform, so that the sound volume level for an arbitrary band is obtained only for the decay waveform output from the first sound source section Aa. Level control for decreasing the amount of attenuation L ′ is performed. In this way, in the third modification shown in FIG. 10, the musical sound waveform in which the synthesis start timing output from the first sound source unit Aa is synthesized later in time with respect to the decay waveform that is earlier in time. In order to be able to reflect the influence of coupled vibration that may occur during synthesis of the attack part, not only the volume level of the decay waveform output from the first sound source part Aa but also the attack output from the second sound source part Ab The amount of attenuation is obtained in consideration of the volume level of the waveform.

(Second embodiment)
In the first embodiment described above and the modification thereof, the musical sound waveform synthesized by the sound source unit A (or the first sound source unit Aa) is passed through the plurality of Notch filters B1 of the filter unit B in order and the harmonic frequencies are close to each other. By reducing the volume level of the overtone within the band for each band, it is possible to synthesize a musical sound waveform that reflects the influence of “coupled vibration”. On the other hand, the second embodiment shown below and its modified examples (hereinafter referred to as the fourth to seventh modified examples for convenience) are not limited to lowering the volume level of overtones for each specific band. Furthermore, it is possible to synthesize a musical sound waveform that more closely reflects the influence of “coupled vibration” by shifting the pitch of multiple overtones that are close in frequency so that they approach a reference frequency that is appropriately determined. It is something that can be done. This will be described below.

  FIG. 11 is a functional block diagram for explaining the outline of the second embodiment of the tone synthesis function. In the second embodiment (referred to as a fourth modified example), the pitch and volume level of the overtones that are arbitrarily designated are changed for each single musical tone waveform corresponding to the notes synthesized in each tone generation channel. As shown in FIG. 11, in this embodiment, a single tone waveform corresponding to a note synthesized by each tone generation channel unit Ac1 of the sound source unit Ac is added to include a tone having a plurality of pitches. Without generating two musical sound waveforms, the single musical sound waveform is output to the channel processing unit T provided corresponding to each sound generation channel unit Ac1 of the sound source unit Ac. Each channel processing unit T has a plurality of harmonic processing units T1 connected in series, and the number thereof is the number of harmonics of notes stored in the harmonic structure database (see FIG. 4) (for example, up to seven eighth harmonics). ). However, since it is rare to perform control to change the pitch or level for all the harmonics of the notes stored in the harmonic structure database, the channel processing unit T has a smaller number (for example, three) of harmonic processing units T1. You may only have.

  Each of the plurality of harmonic processing units T1 connected in series includes one or more arbitrary harmonics from second to eighth harmonics included in a single tone sound waveform output from each sound generation channel Ac1 of the sound source unit Ac. Control is performed to shift the frequency of the harmonic overtone and reduce the volume level by an arbitrary attenuation. The harmonics to be controlled by each harmonic processing unit T1 and the amount of pitch shift or volume level to be controlled are instructed (output) from the determination unit D. Each overtone processing unit T1 includes a band pass filter (BPF) T1a, a pitch shift unit T1b, a level correction unit T1c, an addition unit T1d, and a band eliminate filter (BEF) T1e.

  A single tone tone waveform output from each sound generation channel unit Ac1 of the sound source unit Ac is input to the bandpass filter T1a and the band eliminate filter T1e, and only a specific harmonic (for example, a sixth harmonic) indicated by the determination unit D is obtained. While the included waveform is extracted by the bandpass filter T1a and sent to the pitch shift unit T1b, overtones other than the specific overtone (for example, 2-5th and 7th, 8th overtones other than the 6th overtone) and The waveform including the fundamental tone is extracted by the band eliminate filter T1e and sent to the adding unit T1d. The band pass filter T1a and the band eliminate filter T1e are, for example, IIR filters.

  The specific harmonics extracted by the bandpass filter T1a are harmonics in a band that can cause “coupled vibration” based on the partial sound list and the band list (see FIGS. 3B and 3C). It is specified. More specifically, if the band list has the stored contents as shown in FIG. 3C, the sixth and seventh orders for the channel processing unit T to which a single tone waveform corresponding to the note “A2” is input. A harmonic corresponding to a note “G3” is specified as a processing target by specifying a fifth harmonic and a sixth harmonic as a processing target for the channel processing unit T in which a harmonic overtone is specified as a processing target and a single tone waveform corresponding to the note “C3” is input. The fourth harmonic is designated as a processing target for the channel processing unit T to which the musical sound waveform is input.

  The pitch shift unit T1b changes the frequency of the specific overtone extracted through the bandpass filter T1a based on the pitch shift amount instructed by the determination unit D. It is assumed that the pitch shift amount used when shifting the frequency of the specific overtone is on the frequency axis that the specific overtone will be attracted by the combined vibration from the frequency in the state where the specific overtone does not perform the combined vibration. Calculated as the difference to the point frequency. One hypothetical point that will be attracted by coupled vibrations is that "ceil (b / d)" points are evenly scattered in the range of bandwidth b (Hz), separated by frequency width d (Hz). It is modeled as a thing.

  For example, when three keys to which a note “A2”, a note “C3”, and a note “G3” are assigned are pressed, in the band 3 in which the frequencies of these harmonics are close to each other (FIG. 3B )), The third harmonic of “A2”, the sixth harmonic of “C3”, and the fourth harmonic of “G3” are the middle points of “1540.4 Hz” and “1569.9 Hz” “1555.2 Hz ( Suppose that it is drawn to one hypothetical point at “reference frequency)”. In this case, the seventh harmonic of “A2” is pitch-shifted by “+14.8 Hz” from 1540.4 Hz, and the sixth harmonic of “C3” is pitch-shifted by “−12.9 Hz” from 1568.1 Hz. The frequency of the next overtone is pitch-shifted from 1569.9 Hz by “−14.7 Hz”, and as a result, the frequency of these three overtones is changed to be close to 1555.2 Hz.

  The pitch shift method performed by the pitch shift unit T1b may be any conventionally known method. For example, a pitch shift in a known time domain is preferable. In the time-domain pitch shift, a predetermined window is applied to the waveform of the specific overtone that is input to cut out one frame, and it is resampled at a rate according to the pitch shift amount obtained as described above and compressed in time. Generate a decompressed frame. In this manner, frames that are compressed and expanded one after another while advancing the time position are generated one after another, and a waveform of a specific harmonic that is pitch-shifted is generated by overlapping and joining them.

  The volume level of the specific overtone that has been pitch-shifted by the pitch shift unit T1b is reduced by the level correction unit T1c. In the level correction unit T1c, the volume level of the overtone that has been pitch-shifted is processed according to the original volume level and the attenuation L ′ instructed from the determination unit D. For example, if the attenuation amount L ′ instructed from the determination unit D is the maximum attenuation amount L (for example, −6 dB) stored in the band list, the maximum attenuation amount is closer to “−6 dB” as the volume level of the input overtone increases. On the other hand, the volume level is reduced to a relatively small amount as the volume level of the overtone is reduced.

  Here, a detailed configuration of the level correction unit T1c is shown in FIG. As shown in FIG. 12, the level correcting unit T1c includes an attenuator (Attenator) O, an envelope follower (E.F) P, and a multiplier (Conv) F. The attenuator O decreases the volume level of the input overtone by the attenuation amount L ′. The envelope follower P obtains an instantaneous volume level by envelope-detecting the waveform of the input specific overtone. The multiplier F multiplies the instantaneous volume level from the envelope follower P and the maximum attenuation L instructed from the determination unit D to obtain the attenuation L ′. That is, the attenuation amount L ′ varies up to the maximum value L according to the volume level of the input overtone.

  Returning to the description of FIG. 11, the adding unit T1d is pitch-shifted by the pitch shift unit T1b and the level correcting unit T1c while being pitch-shifted by the waveform including harmonics and fundamentals other than the specific harmonics separated by the band eliminate filter T1e. The waveform of the specific harmonic with the volume level reduced is added to temporarily return to a single tone sound waveform including all the harmonics, and then the single tone sound waveform is sent to the subsequent second harmonic processing unit T1. The succeeding overtone processing unit T1 performs processing related to another overtone different from the overtone processed by the overtone processing unit T1 according to an instruction from the discrimination unit D for the single tone tone waveform output from the preceding overtone processing unit T1. For example, in the case of the channel processing unit T to which a single tone waveform corresponding to the note “A2” is input, the harmonic processing unit T1 processes the sixth harmonic and the harmonic processing unit T2 processes the seventh harmonic. . The harmonic processing unit T1 that is not instructed to process harmonics by the determination unit D may be set to have a flat characteristic or a signal through path (not shown).

  Then, the channel processing unit T outputs to the output adding unit K a single musical tone waveform in which the pitch and volume level are changed with respect to a specific harmonic. The output adding unit K synthesizes one musical tone waveform including a plurality of musical tones corresponding to a plurality of notes by adding the plurality of single tone musical waveforms output from each channel processing unit T. Output as an output signal.

(5th modification)
Various modifications can also be made in the above-described second embodiment in which pitch shift control and volume level control are performed for specific harmonics. FIG. 13 is a functional block diagram showing a first modification (referred to as a fifth modification for convenience) of the second embodiment of the tone synthesis function. In the fifth modified example shown in FIG. 13, a single tone musical sound waveform corresponding to each note is separately divided into a decay waveform and an attack waveform in the tone generator (Ac, Ab), and finally synthesized. By adding the decay waveform and attack waveform of a plurality of notes, one musical sound waveform including a plurality of pitches corresponding to a plurality of notes is generated, and these separately synthesized attack waveforms and decay waveforms are generated. Before adding the waveform, pitch shift control and volume level control are performed on the decay waveform corresponding to each note so as to reflect the influence of “coupled vibration”. Here, the same components as those in the functional block diagrams shown in FIGS. 10 and 11 are denoted by the same reference numerals, and the operations thereof are the same as described above, and thus detailed description thereof is omitted.

  In the fifth modified example shown in FIG. 13, the sound source unit is divided into a first sound source unit Ac that synthesizes a decay waveform and a second sound source unit Ab that synthesizes an attack waveform, and each sound generation channel unit Ab1 of the second sound source unit Ab The output attack waveform is added by the adding unit Ab2 to be an attack waveform including a plurality of pitches corresponding to a plurality of notes, and then sent to the output adding unit K. On the other hand, the decay waveform output from each sound generation channel unit Ac1 of the first sound source unit Ac is sent to the channel processing unit T as it is. As described above, each harmonic processing unit T1 of the channel processing unit T shifts the frequency of a specific harmonic among the second to eighth harmonics included in the decay waveform output from each sound generation channel unit Ac1. In addition to performing pitch shift control, volume level control is performed to reduce the volume level by an arbitrary amount of attenuation. The harmonics to be processed in each harmonic processing unit T1, the pitch shift amount to be changed, and the attenuation amount of the volume level are instructed (output) from the determination unit D. In this way, even when a musical sound waveform that strongly reflects the influence of “coupled vibration” is synthesized by changing the frequency of the specific overtone by performing pitch shift, the decay output from the first sound source unit Ac. While the influence of the coupled vibration is reflected only on the waveform, the attack waveform output from the second sound source unit Ab is not reflected on the attack waveform.

  Here, the tone generation processing for realizing the above-described fourth and fifth modifications will be described, but the “main processing” (see FIG. 5) is not different from the case of the first embodiment. On the other hand, note-on processing (see FIG. 6) and note-off processing (see FIG. 7) are performed differently from those in the first embodiment. FIG. 14 is a flowchart showing a part of note-on processing for realizing the fourth and fifth modified examples of the musical tone synthesis function. However, note-on processing that realizes the fourth and fifth modification examples is the same as the processing shown in FIG. 6 except that the processing in step S16 is different, and the other processing is the same. It is shown.

  Step S16a updates the pitch shift setting of each overtone processing unit T1 in the channel processing unit T in accordance with the update of the partial sound list and the band list accompanying the note addition to the note list. In step S16b, the volume level control setting of each overtone processing unit T1 in the channel processing unit T is updated in accordance with the update of the band list. Thus, in each harmonic processing unit T1 of the channel processing unit T, the second to eighth harmonics included in the single tone waveform (or only the decay waveform) corresponding to the note output from each tone generation channel A1. Among them, pitch shift control for pitch shifting the frequency of the specific overtone is performed, and volume level control for decreasing the volume level by an arbitrary attenuation amount is performed.

  FIG. 15 is a flowchart showing a part of note-off processing for realizing the fourth and fifth modified examples of the tone synthesis function. However, note-off processing that realizes the fourth and fifth modification examples is the same as the processing shown in FIG. 7 except that the processing in step S27 is different, and the other processing is the same. It is shown.

  In step S27a, the pitch shift setting of the harmonic processing unit T1 in the channel processing unit T is updated in accordance with the update of the partial sound list and the band list accompanying the deletion of the note from the note list. Step S27b updates the level control setting of the harmonic processing unit T1 in the channel processing unit T in accordance with the update of the band list. In this case, in each overtone processing unit T1 of the channel processing unit T, a single tone tone waveform (or only a decay waveform) corresponding to other notes other than the note-off notes output from each sound generation channel A1. The pitch shift control is performed to pitch shift the frequency of the specific overtone among the second to eighth harmonics included in the sound, and the volume level control is performed to reduce the volume level by an arbitrary attenuation amount.

(Sixth Modification)
FIG. 16 is a functional block diagram showing a sixth modification of the musical tone synthesis function. In the sixth modification shown in FIG. 16, each sound generation channel unit Ad1 of the sound source unit Ad synthesizes a single tone tone waveform corresponding to a note based on the waveform data read from the waveform memory in response to a sound generation instruction, and generates time-series waves. The high value is sequentially sent to the channel processing unit U. Each channel processing unit U sequentially stores the peak values sent from the sound generation channel unit Ad1 of the sound source unit Ad one by one, and a plurality of tone waveforms formed by the single sound are configured based on the stored peak values. Pitch shift control and volume level control are performed for specific overtones within the band instructed by the determination unit D among overtones. The channel processing unit U includes a channel conversion unit U1, a channel pitch shift unit U2, and a channel level control unit U3.

  The channel conversion unit U1 sequentially extracts time series peak values stored temporarily in a predetermined frame length (for example, 4096 samples), performs a predetermined windowing on this, and performs frequency analysis by Fourier transform (FFT). Convert the complex number to polar coordinates. By performing frequency analysis using this Fourier transform, the tone waveform of a single tone synthesized by the sound source part A is converted to a harmonic frequency component in the vicinity of which the center frequency is an integer multiple of the sampling frequency (44.1 kHz) / frame length (4096). It is decomposed into a number of partial sounds that can be reflected, and the amplitude and phase of each partial sound are obtained (this is called partial sound data). If such a process is repeated while moving the temporal position of the frame to be cut out as time progresses, the temporal transition of the single tone waveform synthesized by each sound generation channel section Ad1 is obtained for each partial sound. A large number of partial sound data obtained in this way are sent simultaneously to the channel pitch shift unit U2.

  The partial sound can reflect, for example, a nearby harmonic component having a center frequency of an integer multiple of sampling frequency (44.1 kHz) / frame length (4096). In other words, the partial sound data obtained as a result of frequency analysis includes a partial sound reflecting around 0 Hz, a partial sound reflecting around 10.8 Hz, a partial sound reflecting around 21.5 Hz, and a partial sound reflecting around 22.05 kHz. Each partial sound may include a harmonic component or may not include a harmonic component. In order to process a specific overtone to reflect the influence of “coupled vibration” (specifically, pitch shift control and volume level control), a partial tone (specifically, a corresponding part) that would reflect the harmonic component is reflected. (Sound data) may be controlled.

  For example, when processing the 7th harmonic of the note “A2”, the frequency of the 7th harmonic of the note “A2” is 1540.4 Hz. Therefore, the center frequency that can include the frequency 1540.4 Hz of this 7th harmonic in the vicinity. Pitch shift control and volume level control may be performed with a partial sound having a frequency of 1539.6 Hz as a control target. At this time, pitch shift control and volume level control may be performed with another partial sound adjacent to the partial sound to be controlled being further controlled. In the case of the above example where the partial sound with the center frequency of 1539.6 Hz is controlled, the partial sound with the center frequency of 1528.8 Hz and the partial sound with the center frequency of 1550.3 Hz correspond to other adjacent partial sounds. .

  The channel pitch shift unit U2 processes the phase of a partial sound that will reflect a specific overtone among a number of partial sounds sent simultaneously from the channel conversion unit U1 in time series in accordance with an instruction from the determination unit D. Pitch shift control is performed. For example, a single tone waveform corresponding to the note “A2” is synthesized in the sound generation channel part Ad1, and the partial sound list and the band list (see FIG. 3B) are obtained from the determination part D in the same manner as in the fourth modification described above. If the instruction to shift the 7th harmonic of the note “A2” from 1540.4 Hz to 1559.3.2 Hz is given based on FIG. 3C, the seventh harmonic of the note “A2” Control is performed to advance the phase of a partial sound having a center frequency of 1539.6 Hz that can include the frequency of 1540.4 Hz by an amount corresponding to the pitch difference. After processing such a phase in a certain frame, with respect to subsequent frames that temporally follow the frame in which the phase has been processed, the next frame that follows so that the phase advances further while the phase continues to the preceding frame. Process phases sequentially.

  The channel level control unit U3 reflects specific overtones among a large number of partial sound data simultaneously transmitted in time series from the channel conversion unit U1 via the channel pitch shift unit U2 according to the instruction of the determination unit D. Volume level control is performed by processing the amplitude of the partial sound. For example, when the tone generation channel unit Ad1 synthesizes a single tone waveform corresponding to the note “A2” and the determination unit D gives an instruction to lower the volume level of the seventh harmonic overtone of the note “A2”. Performs control to reduce the amplitude of the partial sound having the center frequency 1539.6 Hz, which may include the frequency 1540.4 Hz of the seventh harmonic of the note “A2”. At this time, processing is performed to reduce the amplitude by the amount corresponding to the relatively displaced decrease value determined according to the attenuation amount instructed from the determination unit D and the amplitude value of the partial sound as the frequency analysis result. That is, the attenuation amount instructed from the determination unit D is the maximum attenuation amount (for example, −6 dB) stored in the band list, and this maximum attenuation amount is applied uniformly regardless of the amplitude value of the partial sound. Rather than processing the amplitude, the smaller the amplitude of the partial sound, the smaller the amplitude decreases to a maximum of -6 dB, while the smaller the amplitude value of the partial sound, the smaller the amplitude is. To do.

  In the synthesizing unit W for a plurality of channels, a set of a plurality of sound data simultaneously sent in time series from each channel processing unit U provided corresponding to each sound generation channel unit Ad1 of the sound source unit Ad is added. Reconstruct into multiple sound data. The inverse transform unit Y performs inverse Fourier transform (inverse FFT) processing according to a large number of partial sound data including processed partial sound data reconstructed into one set, thereby generating a digital waveform for one frame (4096 samples). Data (for example, PCM data) is generated in time series. The digital waveform data for one frame (4096 samples) is generated one after another as time progresses. By connecting them while shifting in time, multiple pitches corresponding to multiple notes can be used. One musical sound waveform including musical sounds can be output. Note that the inverse transform unit Y is not limited to performing the inverse Fourier transform (inverse FFT) processing as described above, for example, by adding and synthesizing 2048 sine waves generated based on a large number of partial sound data, One musical sound waveform including a plurality of musical tones corresponding to a plurality of notes may be output.

  Here, an example of obtaining the amplitude and phase of each partial sound using a frequency analysis method using Fourier transform processing has been described. However, the present invention is not limited to this, and other known time frequency analysis methods such as a phase vocoder may be used. Good. Alternatively, other time scale analysis by Wevelet may be used.

(Seventh Modification)
In the sixth modification shown in FIG. 16, a single musical tone waveform corresponding to a note once synthesized by each sound generation channel unit Ad1 of the sound source unit Ad is converted into a large number of partial sound data as time frequency information, and these partial sounds are converted. One musical sound waveform including musical tones having a plurality of pitches corresponding to a plurality of notes is synthesized by inversely transforming the sum of the data after processing the data. Instead of synthesizing a single musical tone waveform and generating partial sound data by frequency analysis in each tone generation channel A1 of the sound source part A, a time-series partial sound obtained by converting a single musical tone waveform corresponding to a note into time frequency information in advance. The data may be stored as data, and the stored time-series partial sound data may be sequentially read and output to the channel processing unit T in response to a sound generation instruction. FIG. 17 is a functional block diagram showing a seventh modification of the musical tone synthesis function.

  In the seventh modification shown in FIG. 17, each sound generation channel unit Ae1 of the sound source unit Ae is divided into a plurality of partial sound output units M. Each partial sound output unit M reads out the partial sound data corresponding to the instructed note from the partial sound data which is the time frequency information stored in advance according to the sound generation instruction from the keyboard 4 in time series. It is sent to the channel processing unit U. The partial sound data is the same as the data obtained by the channel conversion unit U1 shown in FIG. 16, and when this partial sound data is read in time series as time progresses, the temporal transition of a single tone tone waveform corresponding to each note is obtained. It has already been explained that it can be obtained for each overtone. The channel pitch shift unit U2 responds to a partial sound that will reflect a specific overtone among a large number of partial sound data transmitted in time series from each partial sound output unit M in accordance with an instruction from the determination unit D. Pitch shift control is performed by processing the phase of the partial sound data to be processed. The channel level control unit U3 reflects specific overtones among a large number of partial sound data simultaneously transmitted in time series from the channel conversion unit U1 via the channel pitch shift unit U2 according to the instruction of the determination unit D. Volume level control is performed by processing the amplitude of partial sound data corresponding to the partial sound that will be. Note that the combining unit W and the inverse converting unit Y for a plurality of channels are the same as in the sixth modified example shown in FIG. 16, and thus the description thereof is omitted here.

In the sixth and seventh modifications described above, digital waveform data is obtained by inversely transforming a set of partial sound data obtained by adding a large number of partial sound data in the synthesis unit W for a plurality of channels. Inverse conversion processing is performed according to the partial sound data output for each channel A1, digital waveform data is generated for each sound generation channel A1, and the plurality of digital waveform data generated for each of the sound generation channels A1 are added. May be.
Also, in the sixth and seventh modified examples described above, as in the fifth modified example shown in FIG. 13, the processing is divided into the decay waveform and the attack waveform, and only the decay waveform is described above. It may be possible to perform pitch shift control and volume level control of a specific overtone by the time frequency analysis.

  Here, the tone generation processing that realizes the sixth and seventh modifications described above will be described. However, the “main processing” (see FIG. 5) is different from the first embodiment and other modifications. Absent. However, note-on processing (see FIG. 6) and note-off processing (see FIG. 7) are performed differently from those in the first embodiment and other modifications. FIG. 18 is a flowchart showing a part of note-on processing for realizing the sixth and seventh modified examples of the musical tone synthesis function. However, note-on processing that realizes the sixth and seventh modification examples is the same as the processing shown in FIG. 6 except that the processing in step S16 is different, and the other processing is the same. Only shown.

  In step S16c, the pitch shift setting to the corresponding partial sound data in the channel processing units U and Ua is updated in accordance with the update of the partial sound list and the band list accompanying the note addition to the note list. In step S16d, the level control setting for the corresponding partial sound data in the channel processing units U and Ua is updated in accordance with the update of the band list. As a result, in each channel pitch shift unit U2 of the channel processing units U and Ua, each of the secondary to eighth orders included in a single tone waveform (or only a decay waveform) corresponding to the note output from each tone generation channel A1. Pitch shift control is performed to pitch shift the frequency of specific overtones up to the above harmonics, and volume level control is performed in the channel level control unit U3 to reduce the volume level of the specific overtones by an arbitrary amount of attenuation.

  FIG. 19 is a flowchart showing a part of note-off processing for realizing the sixth and seventh modified examples of the tone synthesis function. However, note-off processing that realizes the sixth and seventh modified examples is the same as the processing shown in FIG. 7 except that the processing in step S27 is different, and the other processing is the same. Only processing is shown.

  In step S27c, the pitch shift setting to the corresponding partial sound data in the channel processing units U and Ua is updated according to the update of the partial sound list and the band list accompanying the deletion of the note from the note list. In step S27d, the level control setting for the corresponding partial sound data in the channel processing units U and Ua is updated in accordance with the update of the band list. In this case, in each channel pitch shift unit U2 of the channel processing units U and Ua, a single tone waveform (or decay) corresponding to other notes other than the note-off notes output from each sound generation channel A1. Pitch shift control is performed to pitch shift the frequency of the specific harmonics of the second to eighth harmonics included in the waveform only), and the channel level control unit U3 sets the volume level of the specific harmonics by an arbitrary attenuation amount. Decreasing volume level control is performed.

1. CPU, 2. ROM, 3. RAM, 4. Performance operator (keyboard, etc.) 5. Display, 6. Setting operator (switch, etc.) 7. Sound source, 7A. Sound system, 8. Storage device. 9, communication interface, 1D, communication bus (data and address bus), A (Ac, Ad, Ae) / sound source unit, Aa / first sound source unit, Ab / second sound source unit, A1 (Aa1, Ab1, Ac1) , Ad1, Ae1) · Sound generation channel unit, A2 (Aa2, Ab2, J) · Adder unit, B · Filter unit, B1 · Notch filter, C · Control unit, C1 (T1a) · Band pass filter, C2 (P) -Envelope follower, C3 (F)-Multiplier, D-Judgment unit, E-Overtone structure database, G-Input addition unit, K-Output addition unit, M-Partial sound output unit, O-Attenuator, T (U, Ua) Channel processing unit, 1 / overtone processing unit, T1b / pitch shift unit, T1c / level correction unit, T1e / band elimination filter, U1 / channel conversion unit, U2 / channel pitch shift unit, U3 / channel level control unit, W / multiple channels Synthesizer, Y / Inverse Transformer

Claims (4)

A musical sound generating means for simultaneously generating a plurality of different sound waveforms corresponding to notes;
Means for acquiring information indicating the harmonic structure of each of the plurality of musical sound waveforms of different pitches from a harmonic structure database storing the harmonic structure of the musical sound waveform corresponding to each note;
Band specifying means for specifying one or more arbitrary bands including a plurality of overtones having frequencies close to each other among the overtones constituting each of the plurality of generated musical sound waveforms based on the acquired information indicating the overtone structure;
A musical tone synthesizing apparatus comprising level adjusting means for attenuating a volume level of each harmonic contained in each of the specified one to plural bands among the harmonics constituting each of the plurality of musical sound waveforms to be generated.   The level adjustment means attenuates, based on the number of harmonics included in the band, the bandwidth of the band, and the frequency difference between the harmonics whose frequencies are close to each other in the band for each of the specified bands. The musical tone synthesizer according to claim 1, wherein a sound volume level to be determined is determined. The musical sound generating means is configured to generate the plurality of different musical tones having different pitches into a musical tone waveform having an attack portion composed of a non-periodic component and a musical sound waveform having a decay portion composed of a periodic component,
The band specifying means specifies one to a plurality of arbitrary bands including a plurality of overtones whose frequencies are close to each other among the overtones constituting each of the generated sound waveform of the plurality of decay units,
The level adjusting means reduces the volume level of each overtone included in each of the specified one to a plurality of bands among the overtones constituting each of the generated musical tone waveforms of the plurality of decay units. Item 3. The musical tone synthesis apparatus according to Item 1 or 2. On the computer,
A procedure for simultaneously generating a plurality of different tone waveforms corresponding to notes,
A procedure for acquiring information indicating the harmonic structure of each of the plurality of musical sound waveforms having different pitches from a harmonic structure database storing the harmonic structure of the musical sound waveform corresponding to each note;
A procedure for identifying one or more arbitrary bands including a plurality of overtones having frequencies close to each other among the overtones constituting each of the plurality of generated musical sound waveforms based on the acquired information indicating the overtone structure;
The program which performs the procedure which reduces the volume level of each overtone contained in each of the specified 1 thru | or a some zone | band among the overtones which comprise each of the said some musical sound waveform to produce | generate.

Images