JP6402492B2 - Electronic musical instrument, pronunciation control method for electronic musical instrument, and program - Google Patents

Description

  The present invention relates to a sound generation control technique in an electronic musical instrument.

  For example, an electronic musical instrument such as an electronic wind instrument controls a sound source with a sensor that detects a wind performance and a sensed value. In the performance of a wind instrument, a winder simply determines the pitch with a finger and controls the volume and tone by blowing. In an actual wind performance, a wind performer performs by changing the timbre according to the shape of the lips, the position of the tongue, the biting pressure, the shape of the throat, and the like.

  Conventional electronic wind instruments have a small number of sensors and cannot sense the performance technique of these wind instruments. On the other hand, due to the downsizing of sensors and advances in mounting technology, multiple sensors are mounted around the inlet, and the player's blowing pressure, tongue movement, lip position, biting pressure, voice, etc. can be detected. A conventional technique (for example, the technique described in Patent Document 1) that has been made possible is known.

JP 2000-122641 A

  However, the detection value of each sensor does not correspond one-to-one with the performance technique of wind performance. In addition, since the sensors are concentrated in a narrow range of the palate, each sensor detects a value interlocked with each other. Therefore, in the electronic musical instrument, there is a problem that it is difficult to realize a wind performance similar to that of a traditional wind instrument even if the sensor value is directly associated with the volume and tone color.

  An object of this invention is to implement | achieve the desired wind performance similar to a natural musical instrument.

In one example, the acquisition unit that acquires a set of sensor output values output from each of the plurality of types of wind sensors, and the sensor output value from the set of sensor output values acquired by the acquisition unit as a feature amount Generating means for generating a feature vector as an output feature vector, storage means for storing the feature vector based on a specific condition as a representative feature vector, and the output feature vector generated by the generating means, comparing means for calculating the distance between the vectors as compared to the representative feature vector stored in the storage means, based on the distance calculated by said comparing means, said set of sensor output value by the acquisition unit And a control means for controlling the pronunciation at the time of acquisition .

  According to the present invention, it is possible to realize a desired wind performance that is closer to a natural musical instrument.

It is a block diagram which shows the hardware structural example of the electronic musical instrument by embodiment. It is a figure which shows the external example of the electronic musical instrument by embodiment. It is a functional block diagram of an electronic musical instrument. It is a block diagram of a WAVE sound source. It is explanatory drawing of 1st Embodiment of a pronunciation control process. It is a flowchart which shows the example of the whole process of 1st Embodiment of a sound generation control process. It is a flowchart which shows the detailed example of the process in which a musical instrument learns performance. It is a flowchart which shows the detailed example of the process which identifies a learning result. It is a flowchart which shows the detailed example of the process sounded by the identified result. It is a figure which shows the structural example of a presentation apparatus. It is a flowchart which shows the example of the whole process of 2nd Embodiment of a sound generation control process. It is a flowchart which shows the detailed example of the process displayed and pronounced with the identified result. It is explanatory drawing of 3rd Embodiment of a pronunciation control process. It is a flowchart which shows the example of the whole process of 3rd Embodiment of a sound generation control process. It is a flowchart which shows the detailed example of the process which adds physical model modulation | alteration to a sound source. It is a flowchart which shows the detailed example of the process which calculates six link parts. It is a flowchart which shows the detailed example of the process which calculates three mass bodies.

  Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the drawings. In each embodiment of the present invention described below, a musical performance of an electronic musical instrument is digitized by a plurality of sensors, a feature vector having the numeric value as a feature value is generated, and calculation based on the feature vector is performed to perform sound source control Is to do.

  FIG. 1 is a block diagram illustrating an example of a hardware configuration common to each embodiment described below of an electronic musical instrument.

A hardware configuration example shown in FIG. 1 includes a CPU (Central Processing Unit) 101, a ROM (Read Only Memory) 102, a RAM (Random Access Memory) 103, a sensor group 104, and an ADC (Analog Digital Converter). 105, a switch group 106, a GPIO (General Purpose Input Output) 107, a WAVE sound source 108, a DAC (Digital Analog Converter) 109, an acoustic system 110, and a presentation device 111, which are connected to each other via a bus 112. Have. The configuration shown in the figure is an example of a hardware configuration capable of realizing an electronic musical instrument, and such a hardware configuration is not limited to this configuration.
The CPU 101 controls the entire electronic musical instrument. The ROM 102 stores a sound generation control program. The RAM 103 temporarily stores data when the sound generation control program is executed.

  Each output of the sensor group 104 is converted from an analog signal to a digital signal by the ADC 105 and read by the CPU 101.

  The switch group 106 realizes a performance function corresponding to a finger hole of a wind instrument. Each operation state of the switch group 106 is read into the CPU 101 via the GPIO 107.

  A WAVE sound source (wavetable sound source) 108 reads a plurality of musical sound waveform data stored in a ROM built in the WAVE sound source 108 from the switch group 106 at a pitch designated via the GPIO 107 and the CPU 101, and outputs a musical sound signal.

  The musical sound signal output from the WAVE sound source 108 is converted from a digital signal to an analog signal by the DAC 109 and then emitted through the acoustic system 110.

  The presentation device 111 relates to the second embodiment of the sound generation control process, and gives various information to the performer.

  FIG. 2 is a diagram illustrating an example of a casing (outer shape) common to each embodiment described below of the electronic musical instrument. This housing has a main body shape part 1 simulating a wind instrument, a switch part 2 corresponding to a finger hole (corresponding to a part of the switch part 106 in FIG. 1), an opening part 3 that emits an acoustic signal, and a sound signal. Speaker unit 4, various controls 5 other than performance (corresponding to a part of the switch unit 106 in FIG. 1), a plurality of sensors (corresponding to the sensor unit 104 in FIG. 1) and a presentation device (in FIG. The mouthpiece unit 6 is mounted, which corresponds to the presentation device 111.

FIG. 3 is a functional block diagram of the electronic musical instrument shown in FIGS. 1 and 2. This electronic musical instrument includes sound source means 305 (corresponding to the WAVE sound source 108 in FIG. 1) that generates an acoustic signal based on a plurality of parameters. The wind digitizing means 301 (corresponding to the sensor group 104 in FIG. 1) digitizes a plurality of wind performances such as blowing pressure, tongue movement, lip position, biting pressure, voice, etc. A feature vector having the converted signal as a feature amount is generated. The pattern recognition unit 302 (corresponding to the CPU 101, the ROM 102, and the RAM 103 in FIG. 1) stores a representative feature vector corresponding to each performance pattern for each of the plurality of performance patterns related to the wind performance, and performs the numerical value calculation unit 301 at the time of playing. Are compared with the representative feature vectors for each performance pattern, thereby controlling a plurality of parameters in the sound source means 305 based on the comparison result. The presentation unit 303 (corresponding to the presentation device 111 in FIG. 1) calculates the distance between the feature vector generated by the brass digitizing unit 301 at the time of playing and the representative feature vector of the performance pattern serving as a model, and based on the distance. And how much the performance at the time of playing differs from the playing pattern that serves as an example is presented to the playing player. The physical model means 304 vibrates a physical model that schematically represents a vibrating body that can be vibrated by using a plurality of digitized signals output from the sound digitizing means 301 at the time of playing as external inputs, and based on the vibration, a sound source means A plurality of parameters at 305 are controlled.

  Hereinafter, each of the functions shown in FIG. 3 will be described in detail. In order to sense in detail the playing by the performer 306, inside the mouthpiece part 6, the blowing pressure sensor 401, the lip sensor 402 for detecting the position of the upper lip, the bite sensor 403 for detecting the biting pressure of the teeth, the shape of the tongue A plurality of tongue sensors 404 for detecting the sound, a voice sensor 405 for capturing the sound signal of the mouthpiece, and the like are mounted. Each of these sensors 401 to 405 outputs an analog voltage value, and each analog voltage value is converted into a digital voltage value by the ADC 105 of FIG. That is, the sensor group 104 (FIG. 1) of 401 to 405 digitizes a plurality of wind performances and generates a feature vector having the feature values of the plurality of digitized signals as the wind performance digitizing means 301 of FIG. Act as part of

  As another embodiment, a camera that photographs the lips of the blower may be arranged.

  FIG. 4 is a block diagram of the WAVE sound source 108 of FIG. In wind instruments, multiple musical tones are played according to the shape of the palate and the strength of breath. The WAVE sound source 108 includes a wave table 501 (for example, three of # 1 to # 3 in the drawing) each recording a plurality of performance waveforms representative of each performance technique, and is designated by the player's finger arrangement by playing the wind. Each performance waveform is sounded at the same pitch. Each waveform is signal-processed so as to sound in the same phase, and the volume does not cancel unnaturally even if they are superimposed. After that, each multiplier 502 (for example, # 1 to # 3) provided and the adder 503 adjust the pronunciation balance of each performance technique, and then the volume controller 504 adjusts the overall volume if necessary. A musical sound is generated by being sent to the DAC 109 in FIG.

  As another example, a synthesizer or a wave guide sound source may be parameter controlled.

  The first to third embodiments of the sound generation control process of the electronic musical instrument based on the configuration shown in FIGS.

  FIG. 5 is an explanatory diagram of the first embodiment of the sound generation control process. As shown in FIG. 3, there are a plurality of sensors of the mouthpiece unit 6 such as 401 to 405, and outputs from these are sampled at regular intervals by the ADC 105 of FIG. 1 and converted into digital values. If the absolute value of those values or the change value from the previous value is recorded, it becomes a 1 × N feature vector. In FIG. 5, the positions represented by small black squares represent feature vectors on the vector space. In FIG. 5, for simplicity, it is expressed in two dimensions, but in actuality, each point is a multidimensional feature vector having a number of dimensions corresponding to the number of sensors (in the example of FIG. 3, five dimensions 401 to 405). Represent.

  In the learning mode, a plurality of wind performances by the performer are sampled and stored. For example, when the performer changes the shape of the palate for overtone performance, each set of sensor output values corresponding to the performance shape is sampled and a plurality of feature vectors are stored. Simultaneously with the memory, the performer presses a switch that the performance is an overtone performance. Further, the same processing is executed for the tongue playing. A finite number of plane peak values obtained by complementing these feature vectors are detected. The number corresponds to the performance identification number, for example, overtone performance and tongue performance. Those whose distance from the peak is equal to or smaller than a predetermined threshold and those whose distance from the peak is larger than the threshold are identified. As a result, each feature vector is extracted as the inside of the broken lines 601 and 602 in FIG. Each centroid value of these distributions 601 and 602 is calculated and stored as a representative feature vector of the performance technique.

  Next, in the performance mode, each distance between each representative feature vector learned by the electronic musical instrument and a feature vector obtained by the performer's wind performance is calculated. Here, if the sum of the absolute values of the difference between the representative feature vector and the feature vector is a distance, it can be determined that the representative feature vector having the shorter distance is closer to the performance by the performance technique corresponding to the representative feature vector. Each multiplication value inversely proportional to each distance for each representative feature vector is given to each multiplier 502 in the WAVE sound source 108 shown in FIG. 4, and the performance waveforms for each performance technique are mixed to generate a tone signal. .

  FIG. 6 is a flowchart illustrating an example of the overall processing of the first embodiment of the sound generation control processing. This processing is realized as an operation in which the CPU 101 in FIG. 1 executes the sound generation control processing program stored in the ROM 102. This sound generation control processing program realizes the functions of the brass digitizing means 301 and the pattern recognition means 302 of FIG. The sound generation control processing program may be distributed by being recorded on a portable recording medium inserted in a portable recording medium driving device (not shown), or the Internet or a local area network or the like by a communication interface (not shown). It may be possible to obtain from the network. Hereinafter, the configuration of FIG. 1 will be referred to as needed.

  First, the CPU 101 executes processing for the musical instrument to learn to play (step S701). Details of this processing will be described later with reference to FIG.

  Next, the CPU 101 executes processing for identifying a learning result (step S702). Details of this processing will be described later with reference to FIG.

  Next, the CPU 101 executes a process of generating a sound with the identified result (step S703). Details of this processing will be described later with reference to FIG.

  Thereafter, the CPU 101 determines whether or not the performer has pressed the learning mode switch (a part of the switch group 106 in FIG. 1) (mode specifying means) (step S704). If the determination in step S704 is no, the CPU 101 repeatedly executes step S703. As a result of the player pressing the learning mode switch, if the determination in step S704 is YES, the CPU 101 returns to the processing in step S701 and redoes the process in which the instrument learns the performance.

  FIG. 7 is a flowchart showing a detailed example of processing (learning means) in which the musical instrument in step S701 in FIG. 6 learns performance.

  First, the CPU 101 captures each sensor value of the sensor group 104 (corresponding to 401 to 405 in FIG. 3) when the performer performs a musical instrument for learning through the ADC 105, and obtains a sensor output vector value. (Step S801).

  Next, the performer designates which performance waveform the current performance corresponds to by pressing a waveform number switch (a part of the switch group 106 in FIG. 1) (performance pattern designating means), and the CPU 101 is depressed. The waveform number corresponding to the waveform number switch is taken in via the GPIO 107 (step S802).

  The CPU 101 stores the waveform number acquired in step S802 and the sensor output vector value obtained in step S801 as a set in the RAM 103 (step S803).

  The CPU 101 determines whether or not the specified number of times has been completed for all waveform numbers (step S804). If the determination in step S804 is no, the CPU 101 returns to the process in step S801. If the determination in step S804 is YES, the CPU 101 ends the process of the flowchart of FIG. 7 and ends the process of step S701 of FIG.

  The feature vector group indicated by the black square is obtained for each waveform number corresponding to the broken lines 601 and 602 in FIG. 5 by the process in which the musical instrument in step S701 in FIG. 6 illustrated in the flowchart of FIG. Is done.

  FIG. 8 is a flowchart showing a detailed example of the process (learning means) for identifying the learning result in step S702 of FIG.

  First, the CPU 101 initializes a selection number register, which is a variable on the RAM 103, to 0 (zero) (step S901).

  Next, the CPU 101 reads from the RAM 103 the feature vector group corresponding to the waveform number indicated by the selection number register, which is obtained in step S701 in FIG. 6 (step S902).

  Next, the CPU 101 obtains the center of gravity corresponding to the waveform number indicated by the selection number register based on the feature vector group read in step S902, and uses the resulting center of gravity as the representative feature vector corresponding to the waveform number. (Step S903).

  Thereafter, the CPU 101 increments the waveform number indicated by the selection number register by +1 (step S904).

  The CPU 101 determines whether or not the processing has been completed for all the waveforms by determining the waveform number indicated by the selection number register (step S905). If the determination in step S905 is NO, the CPU 101 returns to the process in step S902 and repeatedly executes the representative feature vector calculation process corresponding to the next waveform number. When the determination in step S905 is YES, the CPU 101 ends the process of the flowchart of FIG. 8 and ends the process of identifying the learning result in step S702 of FIG.

  FIG. 9 is a flowchart showing a detailed example of the process of sounding with the identified result in step S703 of FIG.

  First, the CPU 101 takes in each sensor value of the sensor group 104 (corresponding to 401 to 405 in FIG. 3) when the performer performs in the performance mode via the ADC 105, and obtains the feature vector described in FIG. (Step S1001).

  Next, as described in FIG. 5, the CPU 101 determines the distance between the representative feature vector of each waveform number obtained in the RAM 103 in step S702 of FIG. 6 and the feature vector obtained in step S1001. Calculation is performed (step S1002).

  The CPU 101 gives each multiplication value inversely proportional to each distance calculated in step S1002 to each multiplier 502 in FIG. 4 constituting the WAVE sound source 108 in FIG. 1 (step S1003). That is, as the output of the wave table 501 corresponding to a performance technique similar to the current performance, the ratio included in the output musical sound signal increases.

  Next, the CPU 101 acquires operation information of finger operation switches (corresponding to a part of the switch group 106 in FIG. 1) via the GPIO 107, and calculates a pitch designated by the performer according to the finger usage of the wind instrument ( Step S1004).

  The CPU 101 sends the musical tone signal output from the WAVE sound source 108 with the volume determined in step S1003 and the pitch calculated in step S1004 to the DAC 109 for sound generation (step S1005). Thereafter, the CPU 101 ends the process of the flowchart of FIG. 9 and ends the process of sounding with the result identified in step S703 of FIG.

  FIG. 10 is a diagram illustrating a configuration example of the presentation device 111 of FIG. Generally, since a beginner of a wind instrument cannot form a palate for correct performance, the correct learning data may not be recorded in the learning mode described in the first embodiment of the sound generation control process. Therefore, in the second embodiment of the sound generation control process described below, the teacher can set the learning mode as a model and teach the student in the performance mode.

  As shown in FIGS. 10 (a) and 10 (b), the two LEDs, the left LED (light emitting diode) 1101 and the right LED 1102, are visible to the student with the mouthpiece 6 (see FIG. 2) in hand. Placed in. The part represented by the cylinder shown in FIG. 10 is an external view of the LED arrangement.

  In the performance mode, the distance between the feature vector acquired according to the performance of the student and the representative feature vector corresponding to the performance technique that the teacher has learned and specified by the student during the performance is compared. When the distance approaches, the right LED 1102 blinks. When the distance increases, the left LED 1101 blinks. Further, the longer the distance, the faster the blinking, and conversely, the blinking is stopped when the distances match. Students practice the shape of the palate so that the left LED 1101 or the right LED 1102 stops blinking. This practice allows students to take the correct form while changing their palate.

  As another embodiment, the distance may be displayed in 3D on the LCD display of the electronic musical instrument. Further, the distance may be presented by the vibration amount of the vibration vibrator so as to use not only the visual sense but also the tactile sense.

  FIG. 11 is a flowchart illustrating an example of overall processing of the second embodiment of the sound generation control processing using the presentation device 111 described above. This processing is realized as an operation in which the CPU 101 in FIG. 1 executes the sound generation control processing program stored in the ROM 102, similarly to the processing in the first embodiment of the sound generation control processing described in the flowchart of FIG. This sound generation control processing program realizes the functions of the brass digitizing means 301, the pattern recognition means 302, and the presentation means 303 of FIG. This sound generation control processing program may also be distributed by being recorded on a portable recording medium inserted in a portable recording medium driving device (not shown), or the Internet, a local area network, etc. It may be possible to obtain from the network. Hereinafter, the configuration of FIG. 1 will be referred to as needed.

  First, the CPU 101 executes a process for the musical instrument to learn the performance according to the performance of the teacher (step S1201). This process is substantially the same as step S701 in FIG. 6 in the process of the first embodiment of the sound generation control process only when the performer becomes a teacher, and an example of the detailed process is described above. 7 is a flowchart.

  Next, the CPU 101 executes processing for identifying a learning result obtained by the teacher's performance (step S1202). This process is also substantially the same as step S702 of FIG. 6 in the process of the first embodiment of the sound generation control process only when the performer becomes a teacher, and an example of the detailed process is described above. 8 is a flowchart.

  Next, the CPU 101 executes a process of displaying and sounding the identified result (step S1203). Details of this processing will be described later with reference to FIG.

  Then, it shifts to the performance mode by a student. The CPU 101 determines whether or not a learning mode switch (a part of the switch group 106 in FIG. 1) has been pressed (step S1204). If the determination in step S1204 is NO, the CPU 101 repeatedly executes step S1203. If the determination in step S1204 is YES as a result of the teacher pressing the learning mode switch, the CPU 101 returns to the process in step S1201 and redoes the process for the musical instrument to learn the performance by the performance of the teacher.

  FIG. 12 is a flowchart showing a detailed example of the processing to display and pronounce the identified result in step S1203 of FIG.

  First, the CPU 101 inputs a waveform number to be learned based on a student switch input in the switch unit 106 (step S1301).

  Next, the CPU 101 takes in each sensor value of the sensor group 104 (corresponding to 401 to 405 in FIG. 3) when the student performs, through the ADC 105, and obtains the performance vector value described in FIG. S1302). This process is the same as step S1001 of FIG. 9 in the first embodiment of the sound generation control process.

  Next, as described in FIG. 5, the CPU 101 determines the distance between the representative feature vector of each waveform number obtained in the RAM 103 in step S <b> 1202 of FIG. 11 and the feature vector obtained in step S <b> 1301. Calculation is performed (step S1303). This process is also the same as step S1002 of FIG. 9 in the first embodiment of the sound generation control process.

  Next, as described in FIG. 5, the CPU 101, among the representative feature vectors of each waveform number obtained in the RAM 103 in step S <b> 1202 of FIG. 11, the representative feature vector of the waveform number designated by the student in step S <b> 1301, The distance from the feature vector obtained in step S1302 is calculated (step S1303).

  Further, the CPU 101 calculates a difference between the distance calculated in step S1303 during the previous performance and held in the RAM 103 in step S1306, which will be described later, and the distance calculated in step S1303 during the current performance (step S1304).

  As a result of the calculation in step S1304, the CPU 101 determines that the left LED 1101 (see FIG. 10) constituting the presentation device 111 is displayed when the distance calculated this time is larger than the previously calculated distance, and when the distance is decreased. The right LED 1102 (see FIG. 10) constituting the presentation device 111 is blinked at a speed corresponding to the absolute value of the distance calculated in step S1303 (step S1305).

  Thereafter, the CPU 101 stores the distance calculated in step S1303 this time in the RAM 103 (step S1306).

  Finally, the CPU 101 executes a process of generating a sound with the identified result (step S1307). This process is the same as the process in step S703 of FIG. 6 described in the first embodiment of the sound generation control process (the process illustrated in the flowchart of FIG. 9). Thereafter, the CPU 101 ends the process of the flowchart of FIG. 12, and ends the process of displaying and sounding with the identified result in step S1203 of FIG.

  Next, a third embodiment of the sound generation control process will be described. FIG. 13 is an explanatory diagram of a third embodiment of the sound generation control process. When transitioning from a performance waveform detected in the performance mode to another performance waveform, a natural musical instrument does not change smoothly but transits through an unstable performance state in the middle.

  FIG. 13 shows a diagram physical model that schematically shows a vibrating body that realizes the transition state. In this physical model, three mass bodies 1401a, 1401b, 1401c indicated by black circles are mutually connected by three springs (link portions) 1402a, 1402b, 1402c having a braking function indicated by black straight lines. One end is connected to each of the three mass bodies 1401a, 1401b, 1401c, and the other end is fixed and three springs (link portions) 1403a, 1403b, 1403c having a braking function are connected, for a total of six springs. And three mass bodies.

  This physical model is configured to generate self-excited vibrations determined by the position of each mass body, the strength of a spring, and the strength of braking by an external input. In order to operate this, one mass body is moved by the value of one sensor constituting the sensor group 104 to self-excited vibration, the positions of the other two mass bodies that vibrate due to the vibration are extracted, and each performance is obtained. 4, that is, the multiplication value of the multiplier 502 in FIG. 4 may be controlled.

  Specifically, the following calculation is repeatedly executed at a discrete time required for each mass body and each spring.

-Position = force / mass of mass + 2 x position immediately before-position just before the last
... (1)

    ・ Distance = Position−Previous position (2)

・ Force = Spring strength x (Distance-Spring length) + Braking strength x (Distance-Distance just before)
... (3)

  The physical model employed in the present embodiment is a three-mass body free vibration and generally has no analytical solution. Since the self-excited vibration generated by this model is constrained at three locations, its movement can be predicted to some extent, but it moves complicatedly depending on the transition path. This complicated movement that can be predicted to some extent is the performance that determines the performance of the electronic musical instrument. If it is too easy, you will get bored of the performance, and if it is too complicated, you cannot play it. The purpose of this embodiment is a model that mimics the transition state when transitioning from performance to performance. Therefore, it is not always necessary to correspond to the structure of an actual natural musical instrument, and even a model simplified in this way functions sufficiently.

  FIG. 14 is a flowchart showing an example of overall processing of the third embodiment of the sound generation control processing using the above-described physical model. This processing is realized as an operation in which the CPU 101 in FIG. 1 executes the sound generation control processing program stored in the ROM 102, similarly to the processing in the first embodiment of the sound generation control processing described in the flowchart of FIG. This sound generation control processing program realizes the functions of the wind digitizing means 301 and the physical model 304 of FIG. This sound generation control processing program may also be distributed by being recorded on a portable recording medium inserted in a portable recording medium driving device (not shown), or the Internet, a local area network, etc. It may be possible to obtain from the network. Hereinafter, the configuration of FIG. 1 will be referred to as needed.

  First, the CPU 101 executes a process of generating sound based on the identified result (step S1501). This process is the same as the process in step S703 of FIG. 6 described in the first embodiment of the sound generation control process (the process illustrated in the flowchart of FIG. 9). It should be noted that the representative feature vector for each waveform number corresponding to each performance technique is obtained in advance in the RAM 103 by the processing in the learning mode described in the first or second embodiment of the sound generation control processing described above. To do. In this embodiment, the learning mode process is omitted.

  Next, the CPU 101 determines whether or not the maximum volume waveform has changed (step S1502).

  If the determination in step S1502 is NO, the CPU 101 returns to step S1501 and continues the sound generation operation.

  If the determination in step S1502 is YES, the CPU 101 executes processing for adding modulation based on the physical model to the WAVE sound source 108 (FIG. 1) (step S1503).

  FIG. 15 is a flowchart showing a detailed example of the process of adding modulation based on the physical model in step S1503 of FIG. 14 to the sound source.

  First, the CPU 101 captures each sensor value of the current sensor group 104 (corresponding to 401 to 405 in FIG. 3) via the ADC 105, and the sensor value of the mass body 1401a in the physical model illustrated in FIG. The position is changed (step S1601).

  Next, the CPU 101 executes a process of calculating six springs (link portions) in the physical model illustrated in FIG. 13 (step S1602). Details of this processing will be described later using the flowchart of FIG.

  Next, the CPU 101 executes a process of calculating three mass bodies in the physical model illustrated in FIG. 13 (step S1603). Details of this processing will be described later with reference to the flowchart of FIG.

  Finally, the CPU 101 extracts the positions of the mass bodies 1401b and 1401c in the physical model illustrated in FIG. 13 and adds them to the multiplication value of the multiplier 502 in FIG. 4 constituting the WAVE sound source 108 in FIG. 1 (step S1603). ). Thereafter, the CPU 101 ends the process of the flowchart of FIG. 15 and ends the process of step S1503 of FIG.

  FIG. 16 is a flowchart showing a detailed example of processing for calculating six link portions in step S1602 of FIG.

  First, the CPU 101 initializes a counter that is a variable on the RAM 103 for counting the processing of the six springs to 0 (zero) (step S1701).

  Thereafter, the CPU 101 repeatedly executes a series of processes from step S1702 to S1706 six times corresponding to six springs.

  First, the CPU 101 calculates the distance according to the above-described equation (2) (step S1702).

  Next, the CPU 101 calculates “spring strength × (distance−spring length)” in the above-described expression (3) and stores it in the variable L on the RAM 103 (step S1703).

  Next, the CPU 101 calculates “braking strength × (distance−distance immediately before)” in the above-described equation (3) and stores it in the variable M on the RAM 103 (step S1704).

  Further, the CPU 101 calculates the expression (3) by adding the value of the variable L calculated in step S1703 and the value of the variable M calculated in step S1704, and the calculation result corresponds to the currently processed spring. The link force is stored in the link force register on the RAM 103 that holds the link force (step S1705).

  Thereafter, the CPU 101 determines whether or not the calculation has been completed for all the links corresponding to the six springs by determining the value of the counter (step S1706). If the determination in step S1706 is NO, the CPU 101 returns to the process in step S1702 and continues the above calculation for the link corresponding to the next spring. If the determination in step S1706 is YES, the CPU 101 ends the process of the flowchart of FIG. 16, and ends the process of calculating six link units in step S1602 of FIG.

  FIG. 17 is a flowchart showing details of the process of calculating three mass bodies in step S1603 of FIG.

  First, the CPU 101 initializes a counter, which is a variable on the RAM 103 for counting the processing of three mass bodies, to 0 (zero) (step S1801).

  Thereafter, the CPU 101 repeatedly executes a series of processing from steps S1802 to S1805 three times corresponding to three mass bodies.

  First, the CPU 101 sums up the values of the force registers of the links on the RAM 103 corresponding to the links connected to the mass body indicated by the current counter value (step S1802). This register value is calculated in step S1705 in FIG. 16 in step S1602 in FIG. 15 and stored in the RAM 103.

  Next, the CPU 101 calculates a new position using the above-described equation (1) (step S1803).

  The CPU 101 updates the current position of the mass body indicated by the counter value with the new position calculated in step S1803 (step S1804).

  Thereafter, the CPU 101 determines whether or not the calculation has been completed for all three mass bodies by determining the value of the counter (step S1805). If the determination in step S1805 is NO, the CPU 101 returns to the process in step S1802 and continues the above calculation for the next mass body. If the determination in step S1805 is YES, the CPU 101 ends the process of the flowchart of FIG. 17 and ends the process of calculating three mass bodies in step S1603 of FIG.

  As another embodiment related to the third embodiment of the sound generation control process described above, the mass body of the physical model illustrated in FIG. 13 may be changed, or the mass bodies may be further increased. Also, the model may be replaced with the configuration of another physical model.

Regarding the above embodiment, the following additional notes are disclosed.
(Appendix 1)
A feature vector generating means for acquiring a wind power output value from each of the plurality of types of wind sensors and generating a feature vector having the acquired wind output values as feature quantities,
A representative feature vector stored corresponding to a predetermined wind pattern is compared with a feature vector generated by the wind output generating means, and a plurality of musical sounds to be generated by the sound source based on the comparison result are compared. Control means for controlling at least one of the parameters;
An electronic musical instrument characterized by comprising:
(Appendix 2)
The control means stores a plurality of representative feature vectors corresponding to each of a plurality of types of wind patterns,
The sound source includes a plurality of wave tables that store a wind waveform corresponding to each of the plurality of wind patterns, and an output circuit that outputs a musical sound based on an output obtained by adding the outputs of the wave tables.
The control means calculates a distance between the feature vector generated by the feature vector generation means and each of the coordinates as the plurality of representative feature vectors, and is inversely proportional to the distance with respect to each output of each wave table. Add weighted to
The electronic musical instrument according to Supplementary Note 1, wherein
(Appendix 3)
After specifying one wind pattern from the plurality of wind patterns, each time a feature vector is generated by the feature vector generating means, the generated feature vector is stored in association with the specified wind pattern. Repeat the action
For each of the plurality of playing patterns, there is further provided a storage control means for storing the centroid value obtained from the plurality of feature vectors stored in association with each of the playing patterns as a representative feature vector for each performance pattern. The electronic musical instrument according to appendix 1 or 2, characterized by:
(Appendix 4)
And a presentation means for presenting a difference between the wind pattern corresponding to the generated wind output and the previously stored wind pattern to the winder based on the distance calculated by the control means.
The electronic musical instrument according to any one of appendices 1 to 3, characterized in that:
(Appendix 5)
The presenting means is arranged on the left and right of the mouthpiece part for performing the wind, and blinks one of them according to the degree of difference and changes the blinking speed according to the value of the degree of difference. Including light emitting elements,
The electronic musical instrument according to appendix 4, characterized by the above.
(Appendix 6)
Electronic musical instruments
While obtaining a wind output value from each of a plurality of types of wind sensors, generating a feature vector having each of the obtained multiple wind output values as a feature amount,
A representative feature vector stored corresponding to a predetermined playing pattern is compared with the generated feature vector, and at least one of a plurality of parameters of a musical tone to be generated by the sound source is compared based on the comparison result. Sound control method to control.
(Appendix 7)
Obtaining a wind power output value from each of the multiple types of wind sensors, and generating a feature vector having the obtained wind output values as feature quantities,
A representative feature vector stored corresponding to a predetermined playing pattern is compared with the generated feature vector, and at least one of a plurality of parameters of a musical tone to be generated by the sound source is compared based on the comparison result. Controlling step;
A program that causes a computer to execute.

DESCRIPTION OF SYMBOLS 1 Main body shape part 2 Switch part 3 Opening part 4 Speaker part 5 Manipulator 6 Mouthpiece part 101 CPU
102 ROM
103 RAM
104 sensor group 105 ADC
106 Switch group 107 GPIO
108 WAVE sound source 109 DAC
DESCRIPTION OF SYMBOLS 110 Sound system 111 Presentation apparatus 112 Bus 401 Wind pressure sensor 402 Lip sensor 403 Byte sensor 404 Tan sensor 405 Voice sensor 501 Wave table 502 Multiplier 503 Adder 504 Volume adjustment part 1101 Left LED
1102 Right LED

Claims (12)

Acquisition means for acquiring a set of sensor output values output from each of the plurality of types of wind sensors;
Generating means for generating, as an output feature vector, a feature vector having each sensor output value as a feature quantity from a set of sensor output values acquired by the acquiring means;
Storage means for storing the feature vector based on a specific condition as a representative feature vector ;
Comparing means for comparing the output feature vector generated by the generating means with the representative feature vector stored in the storage means to calculate a distance between vectors ;
Control means for controlling sound generation when the set of sensor output values is acquired by the acquisition means based on the distance calculated by the comparison means;
An electronic musical instrument characterized by comprising: The storage means stores the correspondence between the representative feature vector corresponding to a predetermined wind pattern and the wind pattern,
The control means, based on the proximity of the distance between the vectors calculated by the comparison means, a wind pattern by a performance when the set of sensor output values is acquired by the acquisition means is the representative feature vector. Specify whether or not it is a corresponding wind pattern, and control pronunciation based on the specified wind pattern,
The electronic musical instrument according to claim 1. The storage means stores a correspondence relationship between the representative feature vector and waveform data,
The control means specifies waveform data of a musical sound to be generated when the sensor output value set is acquired by the acquisition means based on the distance calculated by the comparison means, and the specified waveform data To control pronunciation based on the
The electronic musical instrument according to claim 1 or 2 , wherein It further comprises setting means for setting one of the learning mode and the performance mode.
When the learning mode is set, the control means stores the output feature vector generated by the generation means as the representative feature vector in the storage means,
When the performance mode is set, the representative feature vector stored in the storage means is used to perform control related to the wind performance.
The electronic musical instrument according to any one of claims 1 to 3 , wherein The control means controls at least one of a plurality of parameters of a musical sound to be generated by a sound source based on a comparison result of the comparison means;
The electronic musical instrument according to any one of claims 1 to 4 , wherein The control means performs control for notifying information based on a comparison result of the comparison means.
An electronic musical instrument according to any one of claims 1 to 5, wherein The storage means is configured to store a plurality of the representative feature vectors corresponding to a plurality of kinds of blowing pattern, stores the blow waveform corresponding to each of said plurality of kinds of blowing pattern,
The comparison means, to the blowing waveform the distance between the target and generated the output feature vector and the plurality of the representative feature vectors by the generation unit vector is calculated respectively, corresponding to each representative feature vector Weighting to be inversely proportional to the distance,
The control means controls sound generation based on an output obtained by adding the plurality of wind waveforms according to the weighting.
The electronic musical instrument according to any one of claims 1 to 6, After specifying one blowing pattern from a plurality kinds of the blowing pattern, each time the output feature vector by the generation unit is generated, the output feature vector said generated by relating corresponding to blowing pattern being said specified Repeat the action to memorize,
For each of the plurality of kinds of blowing pattern, the center of gravity values obtained by the plurality of the feature vectors stored in association with each blowing pattern, the storage control means for storing as said representative feature vector of each of the blowing pattern The electronic musical instrument according to claim 1 , further comprising: And a presentation means for presenting a difference between the wind pattern corresponding to the generated wind output and the previously stored wind pattern to the winder based on the distance between the vectors calculated by the comparing means.
The electronic musical instrument according to any one of claims 1 to 7, The presenting means is arranged on the left and right of the mouthpiece part for performing the wind, and blinks one of them according to the degree of difference and changes the blinking speed according to the value of the degree of difference. Including light emitting elements,
The electronic musical instrument according to claim 9 . Electronic musical instruments
Acquire a set of sensor output values output from each of multiple types of wind sensors,
From the acquired set of sensor output values, a feature vector having each sensor output value as a feature quantity is generated as an output feature vector,
The output feature vector generated by the generation unit is compared with a representative feature vector stored in the storage unit as the feature vector based on a specific condition, and a distance between the vectors is calculated .
A sound generation control method for controlling sound generation when the set of sensor output values is acquired based on the calculated distance . Obtaining a set of sensor output values output from each of a plurality of types of wind sensors;
Generating, as an output feature vector, a feature vector having a feature value of each sensor output value from the set of sensor output values obtained ;
Comparing the output feature vector generated by the generating unit with a representative feature vector stored in a storage unit as the feature vector based on a specific condition, and calculating a distance between vectors ;
Controlling sound generation when the set of sensor output values is acquired based on the calculated distance ;
A program that causes a computer to execute.