JP2006323166A - Musical sound generating device and program thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a musical sound generating device for producing sounds of the same formant in different pitches or producing sounds of the same phrase in different formants while reducing the load of processing operation in a sound source. <P>SOLUTION: In the sound source 10, when mode value Ve is "2", since the waveform (modulated waveform) of each formant element having the pitch corresponding to keying is modulated by an envelope waveform data (modulating waveform) of fixed formant elements, the sound of the same formant is produced in different pitches. When mode value Ve is "3", analysis waveform data (modulated waveform) of fixed formant element read at the original pitch of the original waveform data is modulated by the envelope waveform data (modulating waveform) of each formant element extracted from original waveform data read at a reading speed Vms. Thus, by making the modulating waveform different according to performing operations, the sound of the same phrase is produced in different formant. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a musical sound generating apparatus and a musical sound generating program for generating a vocoder sound.

  Generates a modulated waveform of human voice sound and a modulated waveform of musical instrument sound corresponding to the performance operation, and modulates the formant component extracted from the modulated waveform of human voice sound into the modulated waveform of musical instrument sound, as if the human voice A musical sound generating device that generates such a musical instrument sound (vocoder sound) is known, and this type of device is disclosed in Patent Document 1, for example.

JP 2004-279750 A

  By the way, the device disclosed in Patent Document 1 is configured to perform a frequency analysis in real time to extract a formant component from a modulation waveform of a human voice and to perform a frequency analysis of a modulated waveform of a musical instrument sound in real time. When the processing operation load is high, there is a problem that the same formant cannot be pronounced at different pitches, or the same phrase cannot be pronounced at different formants.

  The present invention has been made in view of such circumstances, and it is possible to generate a musical sound that allows the same formant to be pronounced at different pitches or the same phrase to be pronounced at different formants while reducing the processing operation load on the sound source. An object is to provide a device and a musical tone generation program.

  In order to achieve the above object, according to the first aspect of the present invention, a storage means for storing a modulation waveform for each formant component of a human voice, and a musical sound waveform generated corresponding to a performance operation are converted into a formant component of a human voice. Separately, a modulated waveform forming means for forming a modulated waveform by filtering, and a modulated waveform generated by the modulated waveform generating means by a modulated waveform for each formant component of a human voice read from the storage means in response to a performance operation. And vocoder sound generating means for generating a vocoder sound by modulating the modulation waveform.

  In the invention according to claim 2 subordinate to claim 1, the vocoder sound generating means resets the modulation waveform read from the storage means to the head in response to a key depression from the all key release state. Features.

  According to a third aspect of the present invention, there is provided storage means for storing a modulated waveform obtained by filtering a musical sound waveform for each formant component of a human voice, and a modulation waveform for generating a modulation waveform for each formant component of a human voice corresponding to a performance operation. The modulated waveform filtered by the formant component of the human voice read from the storage means in response to the generating means is modulated by the modulation waveform for each formant component of the human voice generated by the modulation waveform generating means. Vocoder sound generating means for generating vocoder sound is provided.

  The invention according to claim 4 that is dependent on claim 3 is characterized in that the modulation waveform generating means resets the generated modulation waveform to the top in response to a key depression from the all key release state. .

  According to the fifth aspect of the present invention, a modulated waveform forming process for forming a modulated waveform by filtering a musical sound waveform generated corresponding to a performance operation for each formant component of a human voice, and a previously stored human voice The modulated waveform for each formant component of the vocoder is read out in response to the performance operation, and the modulated waveform formed in the modulated waveform generation process is modulated by the read modulated waveform for each formant component of the human voice. A vocoder sound generation process for generating sound is executed by a computer.

  The invention according to claim 6 that is dependent on claim 5 is characterized in that the vocoder sound generation processing resets the modulation waveform to be read out to the top in response to a key depression from an all-key release state.

  The invention according to claim 7 stores a modulation waveform generation process for generating a modulation waveform for each formant component of a human voice in response to a performance operation, and a modulated waveform obtained by filtering a musical sound waveform for each formant component of a human voice. A vocoder sound generation process for generating a vocoder sound by modulating a modulated waveform read in response to a performance operation with a modulation waveform for each formant component of a human voice generated by the modulation waveform generation means by a computer. It is made to perform.

  The invention according to claim 8, which is dependent on claim 7, is characterized in that the modulation waveform generation processing resets the generated modulation waveform to the top in response to a key depression from an all-key release state. .

  According to the first and fifth aspects of the invention, the musical sound waveform generated in response to the performance operation is filtered for each formant component of the human voice to form a modulated waveform, while the pre-stored human voice The modulation waveform for each formant component is read in response to the performance operation, and the vocoder sound is generated by modulating the modulated waveform with the read modulation waveform for each formant component of the human voice, reducing the processing operation load on the sound source The same formant can be pronounced at different pitches.

  According to the second and sixth aspects of the invention, the modulation waveform to be read is reset to the head in response to the key depression from the all key release state, so that the modulation waveform can be reset without performing a special operation. Can do.

  According to the third and seventh aspects of the present invention, a modulated waveform obtained by filtering a musical sound waveform for each formant component of a human voice is stored, and a modulated waveform that is read out in response to a performance operation corresponds to the performance operation. A vocoder sound is generated by modulating with a modulation waveform for each formant component of the human voice generated. Therefore, if the modulation waveform is varied according to the performance operation, the same phrase can be pronounced with different formants while reducing the processing operation load on the sound source.

  According to the fourth and eighth aspects of the present invention, the modulation waveform to be generated is reset to the head in response to the key depression from the all key release state, so that the modulation waveform is reset without performing any special operation. be able to.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
A. Configuration (1) Overall Configuration FIG. 1 is a block diagram showing an overall configuration of an electronic musical instrument according to an embodiment. In this figure, the keyboard 1 generates performance information including key-on / key-off events, key numbers, velocities, and the like in response to pressing / releasing key operations (performance operations). The switch unit 2 includes various switches arranged on the musical instrument panel, and generates a switch event corresponding to the operated switch. As a switch according to the gist of the present invention, the switch unit 2 is provided with a mode switch (not shown) for setting an operation mode. By operating this mode switch, the mode value Ve representing the operation mode is set to any one of “0” to “4”. The contents intended by these mode values Ve will be described later.

  The CPU 3 mainly performs switch processing (described later) according to a switch event generated by the switch unit 2 and keyboard processing (described later) according to performance information supplied from the keyboard 1 to form a musical tone on the sound source 10. Instruct. The processing operation of the CPU 3 according to the gist of the present invention will be described in detail later. The ROM 4 includes a program area and a data area. Various control programs executed by the CPU 3 are stored in the program area of the ROM 4. The various control programs include a main routine, switch processing, and keyboard processing described later. A plurality of types of waveform parameters WP are stored in the data area of the ROM 4. The configuration of the waveform parameter WP will be described later.

  The RAM 5 includes a work area and a data area. Various register / flag data used by the CPU 3 are temporarily stored in the work area of the RAM 5. The configuration of the main register provided in the RAM 5 will be described later. The display unit 6 displays an operation state, a parameter setting state, and the like of each unit of the musical instrument according to a display control signal supplied from the CPU 3. The MIDI interface 7 inputs and outputs external MIDI equipment and MIDI data (MIDI message) in a serial format under the control of the CPU 3. The microphone 8 converts the input voice into a voice signal and outputs it. The A / D converter 9 A / D converts the audio signal input from the microphone 8 and outputs it as input waveform data. The input waveform data output from the A / D converter 9 is stored in, for example, the data area of the RAM 5 under the control of the CPU 3.

  The tone generator 10 follows the waveform generation mode corresponding to the waveform parameter WP supplied from the CPU 3 and the mode value Ve set by the mode switch operation, and the tone parameter (note on / note) generated by the CPU 3 based on the performance information from the keyboard 1. A musical tone is formed according to an off command. The configuration of the sound source 10 will be described later. The waveform ROM 11 stores waveform data corresponding to the 128 types of waveform parameters WP [0] to WP [127] stored in the ROM 4 described above. The sound system 12 performs D / A conversion on the musical sound output out output from the sound source 10 into an analog waveform signal, and performs filtering such as removing unnecessary noise from the D / A converted waveform signal before the signal. Amplify and sound from the speaker.

(2) Configuration of Sound Source 10 Next, the configuration of the sound source 10 will be described with reference to FIG. FIG. 2 is a block diagram showing the configuration of the sound source 10. The sound source 10 shown in this figure is roughly divided into a “waveform generator”, “selector SEL”, “analyzer”, and “synthesizer”.

  The waveform generator is composed of waveform generators OSC1 to OSC32. Each of the waveform generators OSC1 to OSC32 is configured by a well-known waveform memory reading method, and waveform data specified by a waveform parameter WP supplied from the CPU 3 (any of original waveform data, analysis waveform data, and envelope waveform data to be described later) Is read from the waveform ROM 11 to generate a waveform. In these waveform generators OSC1 to OSC32, eight waveform generators are made into one set, and four sets of waveform generators OSC1 to OSC8, OSC9 to OSC16, OSC17 to OSC24, and OSC25 to OSC32 are formed.

  The selection unit SEL includes a plurality of selectors that select each waveform output path of the waveform generators OSC1 to OSC32 according to the mode value Ve supplied from the CPU 3. That is, when the mode value Ve is “0”, the selection unit SEL selects a path for directly inputting the waveform outputs of all the waveform generators OSC1 to OSC32 to the adder K25 of the synthesis unit.

  When the mode value Ve is “1”, the waveform output of the waveform generator OSC1 is input to the bandpass filters BPF M1 to BPF M8 of the analysis unit, and the waveform outputs of the other waveform generators OSC2 to OSC32 are combined. The path to be input to the band pass filters BPF C1 to BPF C8 is selected.

  When the mode value Ve is “2”, the waveform outputs of the waveform generators OSC1 to OSC8 are input to the adders K1 to K8 of the analysis unit, respectively, and the waveform outputs of the other waveform generators OSC9 to OSC32 are combined. The path to be input to the band pass filters BPF C1 to BPF C8 is selected.

  When the mode value Ve is “3”, the waveform output of the waveform generator OSC1 is input to the bandpass filters BPF M1 to BPF M8 of the analysis unit, and the waveform outputs of the waveform generators OSC9 to OSC32 are added by the synthesis unit. The paths to be input to the adders K17 to K24 of the synthesis unit via the units K9 to K16 are selected. In this case, the waveform generators OSC2 to OSC8 are not used.

  When the mode value Ve is “4”, the waveform outputs of the waveform generators OSC1 to OSC8 are input to the adders K1 to K8 of the analysis unit, respectively, and the waveform outputs of the waveform generators OSC9 to OSC32 are added to the synthesis unit. The paths to be input to the adders K17 to K24 of the synthesis unit via the units K9 to K16 are selected.

  The analysis unit is composed of bandpass filters BPF M1 to BPF M8, envelope detectors ENV provided in respective subsequent stages of these bandpass filters BPF M1 to BPF M8, and adders K1 to K8. Each of the bandpass filters BPF M1 to BPF M8 has a pass frequency band corresponding to the formant frequency of the human voice, and extracts a formant component of the waveform input from the waveform generator OSC1.

  The envelope detector ENV includes an absolute value circuit including a rectifier and a smoothing circuit including an LPF, and detects and outputs an envelope level of a formant component extracted by each of the bandpass filters BPF M1 to BPF M8. The adders K1 to K8 add the waveform outputs of the waveform generators OSC1 to OSC8 and the output of the envelope detector ENV and supply them to the multipliers J1 to J8 of the synthesis unit. Each output of the adders K1 to K8 is referred to as a modulation waveform.

  The synthesis unit includes adders K9 to K16, band pass filters BPF C1 to BPF C8, adders K17 to K24, multipliers J1 to J8, and an adder K25. The adder K9 adds the waveform outputs of the waveform generators OSC9, OSC17, and OSC25 and supplies them to the adder K17. The adder K10 adds the waveform outputs of the waveform generators OSC10, OSC18, and OSC26 and supplies them to the adder K18. Similarly, the adder K11 + n (n = 0 to 5) adds the waveform outputs of the waveform generators OSC11 + n, OSC19 + n, and OSC27 + n, and supplies them to the adder K19 + n.

  The bandpass filters BPF C1 to BPF C8 have the same pass frequency band as the bandpass filters BPF M1 to BPF M8 of the analysis unit described above, and are input from the waveform generators OSC2 to OSC32 if the mode value is “1”. If the waveform to be processed is the mode value “2”, the waveform input from the waveform generators OSC9 to OSC32 is divided and filtered into eight pass frequency bands. Adders K17 to K24 add the outputs of adders K9 to K16 and the outputs of bandpass filters BPF C1 to BPF C8. Each output of the adders K17 to K24 is referred to as a modulated waveform. The multipliers J1 to J8 multiply the outputs of the adders K17 to K24 by the outputs of the adders K1 to K8 of the analysis unit described above. The adder 25 adds the outputs of the multipliers J1 to J8 to generate a musical sound output out.

  As described above, the sound source 10 has a configuration for modulating (multiplying) the modulated waveform generated in the synthesizing unit with the modulation waveform output from the analyzing unit. In such a configuration, the modulated waveform output from the analyzing unit and the synthesized waveform are combined. This is characterized in that each waveform content of the modulated waveform generated by the unit is selected in accordance with the mode value Ve supplied from the CPU 3, and the intention of this will be described later.

(3) Configuration of ROM 4 Next, the configuration of the waveform parameter WP stored in the data area of the ROM 4 will be described with reference to FIG. FIG. 3 illustrates an example in which 128 types of waveform parameters WP [0] to WP [127] are stored in the data area of the ROM 4. One waveform parameter WP includes an original waveform parameter Pt0 to Lea0 that represents the read attribute of the original waveform data stored in the waveform ROM 11, and an analysis waveform parameter Sa1 [0] that represents the read attribute of the analysis waveform data stored in the waveform ROM 11. ~ Lea1 [7] and envelope waveform parameters Sa2 [0] ~ Lea2 [7] representing the read attribute of the envelope waveform data stored in the waveform ROM 11.

  The original waveform parameters Pt0 to Lea0 are composed of an original pitch Pt0 representing the pitch of the corresponding original waveform data, a read start address Sa0 of the original waveform data, a loop read start address Lsa0, and a loop read end address Lea0.

  The analysis waveform data is a frequency analysis of the original waveform data in advance using an analysis filter bank (corresponding to the bandpass filters BPF M1 to BPF M8 described above), and for each pass frequency band corresponding to the human formant frequency. This data represents a formant component. That is, the analysis waveform data is waveform data of eight formant components corresponding to the eight pass frequency bands.

  The analysis waveform parameters Sa1 [0] to Lea1 [7] for each formant component include the corresponding analysis waveform data read start addresses Sa1 [0] to Sa1 [7] and loop read start addresses Lsa1 [0] to Lsa1 [7]. ] And loop read end addresses Lea1 [0] to Lea1 [7].

  Envelope waveform data is an analysis filter bank (corresponding to the bandpass filters BPF M1 to BPF M8 described above), and the frequency of the original waveform data is analyzed in advance to correspond to the formant frequency of the human voice. This is waveform data obtained by envelope detection from analysis waveform data representing a formant component. Here, an example of envelope waveform data is shown in FIG. FIG. 4 illustrates envelope waveform data of each formant component assigned to the waveform generators OSC1 to OSC8 (see FIG. 2) of the sound source 10 when the mode value Ve is “2” or “4”.

  The envelope waveform parameters Sa2 [0] to Lea2 [7] of each formant component include the read start addresses Sa2 [0] to Sa2 [7] and the loop read start addresses Lsa2 [0] to Lsa2 [7] of the corresponding envelope waveform data. And loop read end addresses Lea2 [0] to Lea2 [7].

(4) Configuration of RAM 5 Next, the configuration of main registers of the RAM 5 will be described with reference to FIG. In FIG. 5, the mode value Ve set by the mode switch operation is stored in the register Ve. A modulation waveform number is stored in the register Vmp. The modulation waveform number stored in the register Vmp specifies one of the original waveform data, analysis waveform data, and envelope waveform data of human voice. The register Vms stores the modulation waveform reading speed.

  The modulated waveform number is stored in the register Vcp. The modulated waveform number stored in the register Vcp specifies any one of the original waveform data, analysis waveform data, and envelope waveform data of the instrument sound. The counter Kc counts the number of key presses. The register Cos stores the leading number of the waveform generator OSC assigned to generate the modulated waveform. The registers Nn [0] to Nn [31] store key numbers k assigned to the waveform generators OSC1 to OSC32.

  The registers Ca [0] to Ca [31] store current read addresses in the waveform generators OSC1 to OSC32. The registers Lsa [0] to Lsa [31] store the loop read start addresses assigned to the waveform generators OSC1 to OSC32. The registers Lea [0] to Lea [31] store loop read end addresses assigned to the waveform generators OSC1 to OSC32. The registers Pt [0] to Pt [31] store the waveform readout speeds of the waveform generators OSC1 to OSC32.

B. Operation Next, the operation of the embodiment having the above-described configuration will be described with reference to FIGS. In the following, the operation of the main routine executed by the CPU 3 as the overall outline operation will be described first, and then the switch processing and keyboard processing constituting the main routine, and the tone generation timer processing executed by the tone generator 10 at predetermined intervals will be described. The operation will be described.

(1) Operation of Main Routine When the electronic musical instrument having the above configuration is powered on, the CPU 3 executes the main routine shown in FIG. 6 and proceeds to step SA1 to reset various registers and flags stored in the RAM 5. In addition to setting the initial value, an initialization process for instructing the sound source 10 to perform initialization is performed. Then, after the initialization is completed, the process proceeds to step SA2, and a switch process corresponding to the switch operation of the switch unit 2, for example, a process of setting an operation mode according to the mode switch operation is executed.

  Subsequently, in step SA3, a keyboard process for instructing the sound source 10 to generate / mute a musical sound in accordance with the key release operation of the keyboard 1 is executed. In step SA4, other processes such as a process of displaying the contents of the operation mode set in the switch process in step SA2 on the display unit 6 are executed. Thereafter, the above steps SA2 to SA4 are repeatedly executed until the power is turned off.

(1) Operation of Switch Processing Next, the operation of switch processing will be described with reference to FIG. When this process is executed through step SA2 (see FIG. 6) of the main routine described above, the CPU 3 proceeds to step SB1 shown in FIG. 7 and determines whether or not the mode value Ve has been changed by the mode switch operation. To do. If the mode switch operation is not performed and the mode value Ve is not changed, the determination result is “NO”, and the flow proceeds to Step SB7. In step SB7, for example, the modulation waveform number and the modulated waveform number set by operating the timbre parameter selection switch are stored in the register Vmp and the register Vcp, respectively, or the reference pitch set by operating the modulation waveform pitch designation switch. The other processing is performed, such as storing the modulation waveform reading speed corresponding to the above in the register Vms, and this processing is completed.

  On the other hand, when the mode value Ve is changed by the mode switch operation, the determination result in Step SB1 becomes “YES”, and the process proceeds to Step SB2. In step SB2, an all note off event is supplied to the sound source 10, and all OSC mute processing is performed to instruct the waveform generators OSC1 to OSC32 included in the waveform generator of the sound source 10 to stop waveform generation. In step SB2, the value “−1” indicating that waveform generation is stopped is stored in the registers Ca [0] to Ca [31] for temporarily storing the current read addresses of the waveform generators OSC1 to OSC32 in accordance with the all OSC mute processing. Store. Next, in step SB3 and subsequent steps, processing corresponding to the mode value Ve changed according to the mode switch operation is executed. Hereinafter, the operation will be described for each mode value Ve.

<When the mode value Ve is “0”>
As described above, when the mode value Ve is “0”, the waveform outputs of the waveform generators OSC1 to OSC32 (see FIG. 2) of the waveform generation unit in the sound source 10 are output to the adder K25 of the synthesis unit of the sound source 10. This is a form of direct input. In other words, the modulated waveform is not modulated with the modulated waveform to form a vocoder sound, but since all waveform generators OSC1 to OSC32 can be assigned to the instrument sound formation, the process proceeds to step SB4, where the modulated waveform is used. After “0” is stored in the register Cos holding the OSC head number, the above-described step SB7 is executed, and then this process is terminated.

<When Mode Value Ve is “1”>
When the mode value Ve is “1”, the waveform output of the waveform generator OSC1 in the sound source 10 is input to the bandpass filters BPF M1 to BPF M8 of the analysis unit, and the waveform outputs of the other waveform generators OSC2 to OSC32 are input. Is input to the bandpass filters BPF C1 to BPF C8 of the synthesis unit. That is, since the modulated waveform can be assigned to the waveform generators OSC2 to OSC32, the process proceeds to step SB5, and the modulated waveform OSC head number “1” designating the waveform generator OSC2 is stored in the register Cos. After executing step SB7, the present process is terminated.

<When the mode value Ve is “2”>
When the mode value Ve is “2”, the waveform outputs of the waveform generators OSC1 to OSC8 in the sound source 10 are input to the adders K1 to K8 of the analysis unit, respectively, and the waveform outputs of the other waveform generators OSC9 to OSC32 are input. Is input to the bandpass filters BPF C1 to BPF C8 of the synthesis unit. That is, since the modulated waveform can be assigned to the waveform generators OSC9 to OSC32, the process proceeds to step SB6, and the modulated waveform OSC head number “8” designating the waveform generator OSC9 is stored in the register Cos. After executing step SB7, the present process is terminated.

<When the mode value Ve is “3”>
When the mode value Ve is “3”, the waveform output of the waveform generator OSC1 in the sound source 10 is input to the bandpass filters BPF M1 to BPF M8 of the analysis unit, and the waveform outputs of the waveform generators OSC9 to OSC32 are synthesized. In this mode, the signals are respectively input to the adders K17 to K24 of the synthesis unit via the adders K9 to K16 of the unit. That is, since the modulated waveform can be assigned to the waveform generators OSC9 to OSC32, the process proceeds to step SB6, and the modulated waveform OSC head number “8” designating the waveform generator OSC9 is stored in the register Cos. After executing step SB7, the present process is terminated.

<When Mode Value Ve is “4”>
When the mode value Ve is “4”, the waveform outputs of the waveform generators OSC1 to OSC8 in the sound source 10 are input to the adders K1 to K8 of the analysis unit, respectively, and the waveform outputs of the waveform generators OSC9 to OSC32 are synthesized. In this mode, the signals are respectively input to the adders K17 to K24 of the synthesis unit via the adders K9 to K16 of the synthesis unit. That is, since the modulated waveform can be assigned to the waveform generators OSC9 to OSC32, the process proceeds to step SB6, and the modulated waveform OSC head number “8” designating the waveform generator OSC9 is stored in the register Cos. After executing step SB7, the present process is terminated.

(2) Operation of Keyboard Processing Next, the operation of keyboard processing will be described with reference to FIGS. When this processing is executed through step SA3 (see FIG. 6) of the main routine described above, the CPU 3 proceeds to step SC1 shown in FIG. 8 to perform key scanning for detecting the key state of the keyboard 1, and this key. It is determined whether the key state obtained by scanning is “no key change”, “key pressed with key number k”, or “key release with key number k”. If there is no key change, this process is completed without doing anything. If the key number is k, steps SC2 to SC10 are executed. In the case of key release with key number k, steps SC11 to SC15 are executed. Hereinafter, the description will be divided into a case where the key number k is pressed and a case where the key number k is released.

a. In the case of the key depression of the key number k If the key depression is the key number k, the process proceeds to step SC2, and the counter Kc for counting the number of key depressions (hereinafter referred to as the key depression number Kc) is incremented and advanced. Next, in step SC3, it is determined whether or not the stepped key depression number Kc is "1", that is, whether or not the key is depressed from the all key release state. Hereinafter, the operation will be described separately for the case where the key is pressed from the all-key release state and the case where the key is not.

<When the key is pressed from all keys released>
In this case, the determination result in step SC3 is “YES”, the process proceeds to step SC4, and the currently set mode value Ve is determined. If the mode value Ve is “0”, the process proceeds to step SC7 shown in FIG. 9 as in the case where the key is not pressed from the all-key release state (described later).

  If the mode value Ve is “1” or “3”, the process proceeds to step SC5. In step SC5, reading of the original waveform data is started in the waveform parameter WP [Vmp] specified by the modulation waveform number stored in the register Vmp among the waveform parameters WP [0] to WP [127] stored in the ROM 4. Address WP [Vmp]. Sa0 is stored in the register Ca [0], the loop read start address WP [Vmp]. Lsa0 is stored in the register Lsa [0], and the loop read end address WP [Vmp]. After Lea0 is stored in the register Lea [0] and the modulation waveform reading speed stored in the register Vms is stored in the register Pt [0], the process proceeds to Step SC7 described later.

  Thereby, when the mode value Ve is “1” or “3”, the waveform attribute to be assigned to the waveform generator OSC1 of the sound source 10 is set according to the key depression from the all key release state. The waveform generator OSC1 outputs a waveform.

  That is, the waveform generator OSC1 of the sound source 10 reads the read start address WP [Vmp]. Sa0, loop read start address WP [Vmp]. Lsa0 and loop read end address WP [Vmp]. Of register Lea [0]. The original waveform data specified by Lea0 is read from the waveform ROM 11 according to the modulation waveform reading speed of the register Pt [0], and the waveform is output.

  On the other hand, if the mode value Ve is “2” or “4”, the process proceeds to step SC6. In step SC6, the read start address WP [Vmp of the envelope waveform data of each formant component in the waveform parameter WP [Vmp] specified by the modulation waveform number Vmp among the waveform parameters WP [0] to WP [127] of the ROM4. ]. Sa2 [0] to WP [Vmp]. Sa2 [7] is stored in registers Ca [0] to Ca [7], respectively, and loop read start address WP [Vmp]. Lsa2 [0] to WP [Vmp]. Lsa2 [7] is transferred to the registers Lsa [0] to Lsa [7], respectively, and the loop read end address WP [Vmp]. Lea2 [0] to WP [Vmp]. Lea2 [7] is stored in registers Lea [0] to Lea [7], respectively, and the original pitch WP [Vmp]. After storing Pt0 in the registers Pt [0] to Pt [7], the process proceeds to Step SC7 described later.

  Thereby, when the mode value Ve is “2” or “4”, the waveform attribute assigned to the waveform generators OSC1 to OSC8 of the sound source 10 is set according to the key depression from the all key release state. The waveform generators OSC1 to OSC8 output waveforms by referring to the waveform attributes.

  That is, the waveform generators OSC1 to OSC8 of the sound source 10 read the read start addresses WP [Vmp]. Sa2 [0] to WP [Vmp]. Sa2 [7], loop read start address WP [Vmp]. Of registers Lsa [0] to Lsa [7]. Lsa2 [0] to WP [Vmp]. Lsa2 [7] and loop read end address WP [Vmp]. Of registers Lea [0] to Lea [7]. Lea2 [0] to WP [Vmp]. The envelope waveform data of each formant component specified by Lea2 [7] is stored in the original pitch WP [Vmp] .P of the registers Pt [0] to Pt [7]. According to Pt0, the waveform is read from the waveform ROM 11 and output.

<When not pressing from all keys>
In this case, the determination result in step SC3 is “NO”, and the flow proceeds to step SC7 shown in FIG. In step SC7, whether or not the pointer s of the register Ca [s] = − 1 exists within the range of “31” from the modulated waveform OSC head number stored in the register Cos, that is, the target to be stored in the register Cos. The waveform generator OSC (Cos + 1) designated by the modulation waveform OSC head number is searched for the waveform generator OSC to which the sound generation is not assigned from the waveform generator OSC32. If the corresponding waveform generator OSC does not exist, the determination result is “NO”, and this processing is completed. However, if there is a waveform generator OSC that is not assigned to sound generation, the determination result in step SC7 is “YES”. In step SC8, the current mode value Ve is determined.

  If the mode value Ve is any one of “0”, “1”, and “2”, the process proceeds to step SC9. In step SC9, the key number k of the depressed key is stored in the register Nn [s] designated by the pointer s representing the waveform generator OSC (s + 1) not assigned with sound generation. In step SC9, the original waveform data in the waveform parameter WP [Vcp] specified by the modulated waveform number stored in the register Vcp among the waveform parameters WP [0] to WP [127] stored in the ROM 4 is displayed. Read start address WP [Vcp]. Sa0 is stored in the register Ca [s], the loop read start address WP [Vcp]. Lsa0 is stored in the register Lsa [s], and the loop read end address WP [Vcp]. Lea0 is stored in the register Lea [s], and the pitch (modulated waveform reading speed) corresponding to the key number k of the depressed key is stored in the register Pt [s], and this processing is completed.

  As a result, when the mode value Ve is “0”, “1”, or “2”, the waveform generator OSC (s + 1) to which no sound is assigned is searched in response to the key depression, and this waveform generation is performed. A waveform attribute to be assigned to the generator OSC (s + 1) is set, and the waveform generator OSC (s + 1) outputs a waveform with reference to the waveform attribute.

  That is, the waveform generator OSC (s + 1) of the sound source 10 that is not assigned to generate a sound at the time of key depression responds to the key depression by reading the read start address WP [Vcp]. Sa0, loop read start address WP [Vcp]. Of register Lsa [s]. Lsa0 and loop read end address WP [Vcp]. Of register Lea [s]. The original waveform data specified by Lea0 is read from the waveform ROM 11 according to the modulated waveform reading speed of the register Pt [s], and the waveform is output.

  When the mode value Ve is “3” or “4”, the process proceeds to step SC10. In step SC10, the key number k of the depressed key is stored in the registers Nn [s] to Nn [s + 7] designated by the pointer s representing the waveform generator OSC (s + 1) not assigned to sound generation. In step SC10, each formant component in the waveform parameter WP [Vcp] specified by the modulated waveform number stored in the register Vcp among the waveform parameters WP [0] to WP [127] stored in the ROM 4 is used. Read start address WP [Vcp]. Sa1 [s] to WP [Vcp]. Sa1 [s + 7] is stored in registers Ca [s] to Ca [s + 7], and the loop read start address WP [Vcp]. Lsa1 [s] to WP [Vcp]. Lsa1 [s + 7] is stored in the registers Lsa [s] to Lsa [s + 7], and the loop read end address WP [Vcp]. Lea1 [s] to WP [Vcp]. Lea1 [s + 7] is stored in registers Lea [s] to Lea [s + 7], respectively, and the original pitch WP [Vcp]. Pt0 is stored in the registers Pt [s] to Pt [s + 7], and this process ends.

  As a result, when the mode value Ve is “3” or “4”, the waveform generator OSC (s + 1) to which the sound generation is not assigned is searched according to the key depression, and the waveform generator OSC (s + 1) is continuously detected. The waveform attributes to be assigned to the eight waveform generators OSC to be set are set, and the eight waveform generators OSC continuous from the waveform generator OSC (s + 1) each output a waveform with reference to these waveform attributes.

  That is, the eight waveform generators OSC continuous from the waveform generator OSC (s + 1) read the read start addresses WP [Vcp] .C of the registers Ca [s] to Ca [s + 7] according to the key depression. Sa1 [s] to WP [Vcp]. Sa1 [s + 7], loop read start address WP [Vcp]. Of registers Lsa [s] to Lsa [s + 7]. Lsa1 [s] to WP [Vcp]. Lsa1 [s + 7], loop read end address WP [Vcp]. Of registers Lea [s] to Lea [s + 7]. Lea1 [s] to WP [Vcp]. The analysis waveform data of each formant component specified by Lea1 [s + 7] is stored in the original pitch WP [Vcp] .P in the registers Pt [s] to Pt [s + 7]. According to Pt0, the waveform is read from the waveform ROM 11 and output.

b. In the case of key release with key number k If the key release is with key number k, the process proceeds to step SC11 shown in FIG. 8, and the key press number Kc is decremented. Next, at step SC12, whether or not there is a pointer s with the register Nn [s] = k in the range of “31” from the modulated waveform OSC head number stored in the register Cos, that is, stored in the register Cos. The waveform generator OSC (Cos + 1) designated by the modulated waveform OSC head number is searched for the waveform generator OSC to which the key number k of the released key is assigned from the waveform generator OSC32. If there is no corresponding waveform generator OSC, the determination result is “NO”, and this processing is completed. However, if there is a waveform generator OSC to which the key number k of the released key is assigned, The determination result in step SC12 is “YES”, and the flow advances to step SC13 to determine the current mode value Ve.

  If the mode value Ve is any one of “0”, “1”, and “2”, the process proceeds to step SC14. In step SC14, the sound generator 10 is instructed to mute the waveform generator OSC (s + 1) designated by the pointer s and assigned the key number k of the released key, and the waveform generation is stopped in the register Ca [s]. Stores a value “−1” representing “” and ends this processing. As a result, the sound generator 10 stops the waveform generation of the waveform generator OSC (s + 1) to which the key number k of the released key is assigned.

  When the mode value Ve is “3” or “4”, the process proceeds to step SC15. In step SC15, the sound generator 10 is instructed to mute the eight waveform generators OSC (s + 1) to OSC (s + 8) designated by the pointers s to s + 7 and assigned the key number k of the released key. A value “−1” indicating that waveform generation is stopped is stored in Ca [s] to Ca [s + 7], and this processing is completed. As a result, the sound source 10 stops the waveform generation of the eight waveform generators OSC (s + 1) to OSC (s + 8) to which the key number k of the released key is assigned.

(3) Operation of Sound Generation Timer Process Next, the operation of the sound generation timer process executed by the sound source 10 will be described with reference to FIGS. The sound source 10 executes this process at predetermined intervals by a timer interrupt. When the execution timing comes, the process proceeds to step SD1 shown in FIG. In step SD1, the registers Ad, Act, Amt, Ac [0] to Ac [7], Am [0] to Am [7] and the pointer n are reset to zero.

  The register Ad holds a waveform value that is directly input to the adder K25 (see FIG. 2) of the synthesis unit of the sound source 10. The register Act holds waveform values input to the bandpass filters BPF C1 to BPF C8 (see FIG. 2) of the synthesis unit of the sound source 10. The register Amt holds waveform values input to the bandpass filters BPF M1 to BPF M8 (see FIG. 2) of the analysis unit of the sound source 10. The registers Ac [0] to Ac [7] hold waveform values respectively input to the adders K17 to K24 (see FIG. 2) of the synthesis unit of the sound source unit 10. The registers Am [0] to Am [7] hold waveform values respectively input to the adders K1 to K8 (see FIG. 2) of the analysis unit of the sound source unit 10. The pointer n designates the waveform generators OSC1 to OSC32 of the sound source 10.

  Next, in step SD2, whether or not the value of the register Ca [n] specified by the pointer n is a value “−1” indicating that waveform generation is stopped, that is, the waveform generator OSC (n + 1) specified by the pointer n. ) Is being muted. If the waveform generator OSC (n + 1) is muted, the determination result is “YES”, the process proceeds to step SD11, and the pointer n is incremented to be incremented. In step SD12, it is determined whether or not the value of the incremented pointer n has reached “32”, that is, whether or not all the waveform generators OSC1 to OSC32 have been muted. If the check has not been completed, the determination result is “NO”, and the process returns to step SD2.

  In step SD2, when the waveform generator OSC that is generating sound (generating a waveform) is found, the determination result in step SD2 is “NO”, and the process proceeds to step SD3. In step SD3, the waveform read speed of the register Pt [n] is added to the register Ca [n] designated by the pointer n, and the current read address of the register Ca [n] is updated. In step SD3, when the current read address Ca [n] obtained by adding the waveform read speed Pt [n] exceeds the loop read end address Lea [n], the loop read ends from the current read address Ca [n]. The loop read start address Lsa [n] is added to the value obtained by subtracting the address Lea [n], and this is updated as the current read address Ca [n].

  Next, in step SD4, the waveform data is read from the waveform ROM 11 in accordance with the current read address Ca [n] updated in step SD3, and the read waveform data is subjected to interpolation processing to generate a waveform value d. In addition, the interpolation process said here refers to the process which interpolates the waveform data read by the integer part address by the decimal part address. Thus, when a new waveform value d is generated in the waveform generator OSC (n + 1) designated by the pointer n, steps SD5 and subsequent steps are executed to form a waveform according to the mode value Ve. The waveform forming operation for each mode value Ve will be described below.

<When the mode value Ve is “0”>
In this case, the process proceeds to step SD6 through step SD5, and after adding the waveform value d generated in step SD4 to the register Ad, the process proceeds to step SD11.
Therefore, if the mode value Ve is “0”, the waveform outputs of all the waveform generators OSC1 to OSC32 are accumulated in the register Ad.

<When Mode Value Ve is “1”>
In this case, the process proceeds to step SD7 via step SD5. In step SD7, if the value of the pointer n is “0”, the waveform value d generated by the waveform generator OSC1 is stored in the register Amt, and if the value of the pointer n is larger than “0”, the waveform generator OSC2 is set. After the waveform value d generated by the OSC 32 is added to the register Act, the process proceeds to step SD11 described above.
Therefore, when the mode value Ve is “1”, the waveform output of the waveform generator OSC1 is held in the register Amt, and the waveform outputs of the waveform generators OSC2 to OSC32 are accumulated in the register Act.

<When the mode value Ve is “2”>
In this case, the process proceeds to step SD8 via step SD5. In step SD8, if the value of the pointer n is “0” to “7”, the waveform values generated by the waveform generators OSC1 to OSC8 in the registers Am [0] to Am [7] specified by the pointer n. d is stored, and when the value of the pointer n is “8” or more, the waveform values d generated by the waveform generators OSC9 to OSC32 are added to the register Act, and then the process proceeds to the above-described step SD11.
Therefore, if the mode value Ve is “2”, the waveform outputs of the waveform generators OSC1 to OSC8 are held in the registers Am [0] to Am [7], and the waveform outputs of the waveform generators OSC9 to OSC32 are registered in the register. Accumulated with Act.

<When the mode value Ve is “3”>
In this case, the process proceeds to step SD9 via step SD5. In step SD9, if the value of the pointer n is “0”, the waveform value d generated by the waveform generator OSC1 is stored in the register Amt. In step SD9, if the value of the pointer n is “8” or more, the waveform generators OSC9 to OSC9 corresponding to the registers Ac [0] to Ac [7] specified by the remainder of the pointer “8” are stored. The waveform value d generated by the OSC 32 is added. That is, the register Ac [0] adds the waveform values d of the waveform generators OSC9, OSC17, and OSC25, and the register Ac [1] adds the waveform values d of the waveform generators OSC10, OSC18, and OSC26. Thereafter, similarly, the register Ac [7] adds the waveform values d of the waveform generators OSC16, OSC24, and OSC32.
Therefore, if the mode value Ve is “3”, the waveform output of the waveform generator OSC1 is held in the register Amt and is input to the adders K17 to K24 (see FIG. 2) of the synthesis unit, respectively. The waveform output of the OSC 32 is added in the registers Ac [0] to Ac [7].

<When Mode Value Ve is “4”>
In this case, the process proceeds to step SD10 via step SD5. In step SD10, if the value of the pointer n is “0” to “7”, the waveform values generated by the waveform generators OSC1 to OSC8 in the registers Am [0] to Am [7] specified by the pointer n. Each d is stored. In step SD10, if the value of the pointer n is “8” or more, the waveform generators OSC9 to OSC9 corresponding to the registers Ac [0] to Ac [7] specified by the remainder of the pointer “8” are stored. The waveform value d generated by the OSC 32 is added. That is, the register Ac [0] adds the waveform values d of the waveform generators OSC9, OSC17 and OSC25, and the register Ac [1] adds the waveform values d of the waveform generators OSC10, OSC18 and OSC26. Thereafter, similarly, the register Ac [7] adds the waveform values d of the waveform generators OSC16, OSC24, and OSC32.
Therefore, if the mode value Ve is “4”, the waveform outputs of the waveform generators OSC1 to OSC8 are held in the registers Am [0] to Am [7], and adders K17 to K24 (see FIG. 2) of the synthesis unit. ), The waveform outputs of the waveform generators OSC9 to OSC32 respectively input to the register Ac [0] to Ac [7].

  As described above, when the waveform generation of the waveform generators OSC1 to OSC32 according to the mode value Ve is completed, the determination result in step SD12 is “YES”, and the sound source 10 proceeds to step SD13 illustrated in FIG. After step SD13, the waveform is output according to the mode value Ve. The waveform output operation for each mode value Ve will be described below.

<When the mode value Ve is “0”>
In this case, the process proceeds to step SD14 via step SD13. In step SD14, the waveform outputs of the waveform generators OSC1 to OSC32 accumulated in the register Ad are stored in the register Output, and in the subsequent step SD15, the contents of the register Output are output as the musical sound output out.

<When Mode Value Ve is “1”>
In this case, the process proceeds to step SD16 via step SD13. In step SD16, the waveform output of the waveform generator OSC1 stored in the register Amt is input to the bandpass filters BPF M1 to BPF M8 (see FIG. 2) of the analysis unit, and each of the bandpass filters BPF M1 to BPF M8 is further input. Registers Am [0] to Am [7] are used as waveforms to be inputted to the adders K1 to K8 (see FIG. 2) of the analysis unit, respectively, of the envelope waveform of the formant component extracted by the envelope detector ENV provided in the subsequent stage. To store.

  In step SD16, each waveform output of the waveform generators OSC2 to OSC32 accumulated in the register Act is input to the bandpass filters BPF C1 to BPF C8 (see FIG. 2) of the synthesis unit and extracted. The formant components are stored in the registers Ac [0] to Ac [7] as waveforms respectively input to the adders K17 to K24 (see FIG. 2) of the synthesis unit.

  Next, in step SD17, in the multipliers J1 to J8 (see FIG. 2) of the synthesizer, the outputs (modulated waveforms) of the adders K1 to K8 of the analyzer are output (covered). The modulated waveform data is multiplied (modulated) to generate modulated waveform data, which is stored in the register Output. Then, the process proceeds to step SD15, and the modulated waveform data stored in the register Output is output as the musical sound output out.

  In step SD17, if the mode value Ve is “1”, the modulation waveform is stored in the registers Am [0] to Am [7], and is the envelope waveform of each formant component extracted from the waveform output of the waveform generator OSC1. is there. On the other hand, the modulated waveform is a waveform of each formant component stored in the registers Ac [0] to Ac [7] and extracted from each waveform accumulated value of the waveform generators OSC2 to OSC32.

<When the mode value Ve is “2”>
In this case, the process proceeds to step SD18 via step SD13. In step SD18, the waveform outputs of the waveform generators OSC9 to OSC32 accumulated in the register Act are input to the bandpass filters BPF C1 to BPF C8 (see FIG. 2) of the synthesis unit and extracted. Are stored in the registers Ac [0] to Ac [7] as waveforms respectively input to the adders K17 to K24 (see FIG. 2) of the synthesis unit.

  In step SD17, the outputs (modulated waveforms) of the adders K17 to K24 of the synthesizer are output from the multipliers J1 to J8 of the synthesizer by the outputs (modulated waveforms) of the adders K1 to K8 of the analyzer. After being multiplied (modulated) to generate modulated waveform data and stored in the register Output, the process proceeds to step SD15, and the modulated waveform data stored in the register Output is output as a musical sound output out.

  In step SD17, if the mode value Ve is “2”, the modulation waveform is stored in the registers Am [0] to Am [7] in the above-described step SD8 (see FIG. 10), and the waveform generators OSC1 to OSC8 Each waveform output. On the other hand, the modulated waveform is the waveform of each formant component stored in the registers Ac [0] to Ac [7] and extracted from the waveform accumulated values of the waveform generators OSC9 to OSC32.

<When the mode value Ve is “3”>
In this case, the process proceeds to step SD19 via step SD13. In step SD19, the waveform output of the waveform generator OSC1 stored in the register Amt is input to the bandpass filters BPF M1 to BPF M8 (see FIG. 2) of the analyzer, and each of the bandpass filters BPF M1 to BPF M8 is further input. Registers Am [0] to Am [0] are used as waveforms to be inputted to the adders K1 to K8 (see FIG. 2) of the analysis unit, respectively, of the envelope level (envelope waveform) of the formant component extracted by the envelope detector ENV provided in the subsequent stage. Store in Am [7].

  Next, in step SD17, the outputs (modulated waveforms) of the adders K17 to K24 of the synthesizer are multiplied by the outputs (modulated waveforms) of the adders K1 to K8 of the analyzer in the multipliers J1 to J8 of the synthesizer. After (modulation), modulated waveform data is generated and stored in the register Output, the process proceeds to step SD15, and the modulated waveform data stored in the register Output is output as the musical sound output out.

  In step SD17, if the mode value Ve is “3”, the modulation waveform is stored in the registers Am [0] to Am [7], and is the envelope waveform of each formant component extracted from the waveform output of the waveform generator OSC1. is there. On the other hand, the modulated waveform is stored in the registers Ac [0] to Ac [7] in the above-described step SD9 (see FIG. 10), and is generated into the adders K17 to K24 (see FIG. 2) of the synthesizer. This is a waveform output of the devices OSC9 to OSC32.

<When Mode Value Ve is “4”>
In this case, the process proceeds to step SD17 via step SD13. In step SD17, the multipliers J1 to J8 of the synthesizer multiply (modulate) the outputs (modulated waveforms) of the adders K17 to K24 of the synthesizer by the outputs (modulated waveforms) of the adders K1 to K8 of the analyzer. ) To generate post-modulation waveform data and store it in the register Output, and then proceed to Step SD15 to output the post-modulation waveform data stored in the register Output as a musical sound output out.

  If the mode value Ve is “4” in step SD17, the modulation waveform is stored in the registers Am [0] to Am [7] in step SD10 (see FIG. 10) described above, and the waveform generators OSC1 to OSC8 Each waveform output. On the other hand, the modulated waveforms are waveform outputs of waveform generators OSC9 to OSC32 that are stored in the registers Ac [0] to Ac [7] and input to the adders K17 to K24 (see FIG. 2) of the synthesis unit, respectively.

  As described above, in the present embodiment, when the mode value Ve is “2”, the envelope waveform data (modulation waveform) of the fixed formant components generated by the waveform generators OSC1 to OSC8 at a constant reading speed. ) Modulates the waveform of each formant component (modulated waveform) extracted from the accumulated value of each waveform generated by the waveform generators OSC9 to OSC32 at a pitch according to the key depression, so that the same formant is generated at a different pitch. Can be made. In addition, when the mode value Ve is “2”, since the real-time frequency analysis using the bandpass filters BPF M1 to BPF M8 of the sound source 10 is not performed, the processing operation load on the sound source 10 may be reduced. it can.

  When the mode value Ve is “3”, the original waveform data is obtained from the envelope waveform data (modulated waveform) of each formant component extracted from the original waveform data generated by the waveform generator OSC1 at the preset read speed Vms. The analysis waveform data (modulated waveform) of the fixed formant component generated by the waveform generators OSC9 to OSC32 is modulated at the original pitch. In the above-described embodiment, the reading speed Vms is fixed to simplify the description. However, for example, the reading speed Vms is changed according to the key number k of the key to be pressed, or the key pressing is performed. If the modulation waveform number Vmp is changed according to the key to be played and the modulation waveform is varied according to the performance operation, the same phrase can be pronounced with different formants. In addition, when the mode value Ve is “3”, since the real-time frequency analysis using the bandpass filters BPF C1 to BPF C8 of the sound source 10 is not performed, the processing operation load on the sound source 10 may be reduced. it can.

  In the above-described embodiment, when the mode value Ve is “2” or “4”, the envelope waveform data of each formant component stored in the waveform ROM 11 is read and used as a modulation waveform. Alternatively, the analysis waveform data of each formant component may be read from the waveform ROM 11 and input to the envelope detector ENV of the analysis unit, thereby forming a modulation waveform composed of the envelope waveform data of each formant component. In this way, it is not necessary to store the envelope waveform data of each formant component in the waveform ROM 11, so that the memory capacity of the waveform ROM 11 can be reduced.

  In the present embodiment, the mode value Ve is selected by a user operation. Instead of this, the CPU 3 automatically selects the mode value Ve according to the load of the sound source 10, for example, a high number of simultaneous pronunciations. It is also possible to select a mode value Ve “4” that does not perform frequency analysis for extracting formant components in real time at the time of loading, and to select a mode value “1” when the load is low.

  Further, when the analysis waveform data is used for modulation, the eight waveform generators OSC are always occupied regardless of the amplitude value of each formant component of the modulation waveform. For the formant component having a small value, the waveform generation of the corresponding waveform generator OSC may be stopped. In this way, since the occupation rate of the waveform generator OSC is reduced, the number of simultaneous sounds can be increased.

  In addition, in the present embodiment, the waveform ROM 11 stores the original waveform data, the analysis waveform data, and the envelope waveform data in advance. However, the present invention is not limited thereto, and the microphone 8 and the A / D converter 9 are used. The sampled original waveform data is stored in the data area of the RAM 5, and the analysis waveform data and envelope waveform data are generated by performing frequency analysis and envelope detection on the original waveform data stored in the RAM 5 using the analysis unit of the sound source 10. It is also possible to save in the data area of the RAM 5.

It is a block diagram which shows the whole structure of one Embodiment by this invention. 2 is a block diagram showing a configuration of a sound source 10. FIG. It is a figure which shows the structure of the waveform parameter WP memorize | stored in ROM4. It is a figure which shows an example of the envelope waveform data of each formant component allocated to waveform generator OSC1-OSC8. 3 is a diagram illustrating a main register configuration provided in a RAM 5. FIG. It is a flowchart which shows operation | movement of a main routine. It is a flowchart which shows the operation | movement of a switch process. It is a flowchart which shows the operation | movement of a keyboard process. It is a flowchart which shows the operation | movement of a keyboard process. It is a flowchart which shows the operation | movement of a sound generation timer process. It is a flowchart which shows the operation | movement of a sound generation timer process.

Explanation of symbols

1 Keyboard 2 Switch 3 CPU
4 ROM
5 RAM
6 Display unit 7 MIDI interface 8 Microphone 9 A / D converter 10 Sound source 11 Waveform ROM
12 Sound system

Claims (8)

Storage means for storing a modulation waveform for each formant component of human voice;
A modulated waveform forming means for filtering a musical sound waveform generated corresponding to a performance operation according to a formant component of a human voice to form a modulated waveform;
Vocoder sound generating means for generating a vocoder sound by modulating the modulated waveform formed by the modulated waveform generating means according to a modulation waveform for each formant component of a human voice read from the storage means in response to a performance operation; A musical sound generator characterized by comprising. 2. A musical sound generating apparatus according to claim 1, wherein said vocoder sound generating means resets a modulation waveform read from said storage means to the head in response to a key depression from a full key release state. Storage means for storing a modulated waveform obtained by filtering a musical sound waveform for each formant component of a human voice;
Modulation waveform generating means for generating a modulation waveform for each formant component of human voice in response to a performance operation;
A vocoder sound is generated by modulating a modulated waveform filtered for each formant component of a human voice read from the storage means in response to a performance operation, with a modulation waveform for each formant component of the human voice generated by the modulation waveform generating means. And a vocoder sound generating means. 4. A musical tone generating apparatus according to claim 3, wherein said modulation waveform generating means resets the generated modulation waveform to the head in response to a key depression from an all key release state. A modulated waveform forming process for forming a modulated waveform by filtering a musical sound waveform generated corresponding to a performance operation according to a formant component of a human voice;
A modulated waveform for each formant component of a human voice stored in advance is read in response to a performance operation, and a modulated waveform formed by the modulated waveform generation process is read out according to the read modulation waveform for each formant component of a human voice. A musical sound generating program characterized by causing a computer to execute vocoder sound generation processing for generating a vocoder sound by modulating a waveform. 6. The musical tone generation program according to claim 5, wherein the vocoder sound generation process resets the modulation waveform to be read out to the head in response to a key depression from the all key release state. A modulation waveform generation process for generating a modulation waveform for each formant component of human voice in response to a performance operation;
A modulated waveform obtained by filtering a musical sound waveform according to a formant component of a human voice is stored, and a modulated waveform read out in response to a performance operation is converted into a modulated waveform for each formant component of a human voice generated by the modulation waveform generating means. A musical sound generating program that causes a vocoder sound generating process to generate a vocoder sound by modulating with a computer. 8. The musical tone generation program according to claim 7, wherein the modulation waveform generation processing resets the generated modulation waveform to the head in response to a key depression from the all key release state.

Images