JP2800465B2 - Electronic musical instrument - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electronic musical instrument for playing lyrics by human-generated voice (human voice) as a musical tone, and more particularly to an electronic musical instrument capable of performing double-tone performance by human voice.

[0002]

2. Description of the Related Art Japanese Patent Publication No. 59-8838 discloses an electronic musical instrument capable of generating a human voice by giving musical tone formation information. An electronic musical instrument has been disclosed which reads out and plays the lyrics at the pitch of the key depression, thereby playing the lyrics by the human voice of the electronic musical instrument.

[0003] In today's electronic musical instruments, simultaneous generation of a plurality of musical tones, that is, performance with multiple tones, is the mainstream.
However, with the electronic musical instrument disclosed in the above publication, simultaneous generation of a plurality of human voices, that is, performance of a voice chorus or the like cannot be realized.

[0004]

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems in the prior art, and it is an object of the present invention to provide an electronic musical instrument capable of performing a multi-tone performance in which a voice is a musical tone.

[0005]

In order to achieve the above object, an electronic musical instrument according to the present invention comprises a pitch designation information generating means for generating pitch designation information, and a plurality of formants characterizing the pronunciation of human voice sounds. when the formant information storage means for storing sequence information, in response to the occurrence of pitch designation information from the previous Kion high specification information generating means, a predetermined sequence of
In met sequentially read reading means to the time-series information
The pitch designation information is continuously generated within a predetermined time.
Other than the first consecutive pitch designation information
And which do not proceed with the shea Sequence So every each pitch designation information generated from said pitch designation information generation means, a sound based on the sequential formant information is read by the sound high specification information and said reading means A formant forming sound source to be formed, which is capable of forming a plurality of sounds in parallel.

In one embodiment of the present invention, there is further provided performance operation information input means for inputting performance operation information other than the pitch designation information, and the reading means reads formant information from the formant information storage means. The reading point at that time is changed according to the performance operation information.

The pitch designation information generating means includes a keyboard (keyboard), a MIDI interface, and a card, tape, disk, RO capable of reading melody information.
A musical score memory such as M and RAM can be used.

The voice designation information input means includes a syllable designation switch group such as a keyboard such as a word processor.
Alternatively, a memory unit capable of reading syllable information such as lyrics can be used.

As a formant-forming sound source, a formant sound synthesizing apparatus disclosed by the present applicant in Japanese Patent Application Laid-Open No. 2-262698 can be used.

As the performance operation information input means, there are performance operators such as a key velocity detection means, a modulation wheel, and a pedal.

[0011]

The voice (human voice) generated by a human is composed of a plurality of formants, and these formants each fluctuate with time.

According to the present invention, time series information (formant locus) of one phoneme or syllable (50 tones, each voiced sound, semi-voiced sound, etc.) or a plurality of formants characterizing each syllable is stored. Given it,
The data is read out according to the sequence and supplied to the formant forming sound source. A formant forming sound source sequentially supplies formant information, such as human voice sound characterized by the plurality of formants according to the formant frequency and formant level information and the pitch designation information supplied from the pitch designation information generating means. To form This allows
A human voice having a pitch specified by the pitch specification information generating means can be synthesized. Furthermore, when a plurality of pitches are specified at the same time, the formant-forming sound source generates human voices of the plurality of pitches simultaneously in parallel. In addition, predetermined time
When a plurality of pitch designation information occurs consecutively in the
The reading means reads the first piece of information
Depending on the remaining information,
Does not advance the sequence. In this case, multiple pitch designations
Time series information of the same phoneme or the same syllable
Is spilled out. This allows multiple pitches and the same voice to be played simultaneously.
The sound, that is, the chorus singing, becomes possible.

[0013]

Therefore, according to the electronic musical instrument of the present invention, it is possible to realize a double tone performance (chorus) by voice.

For example, by giving the same musical tone formation information (formant information) to a plurality of depressed keys that are depressed almost simultaneously in time, a plurality of musical tones (human voices) are generated at the same syllable and at multiple pitches. be able to.

Furthermore, if the performance operator information such as key velocity and modulation wheel to change the reading point when reading or et formant information formant information storage hands stage, is possible to realize a more dynamic change in tone it can.

[0016]

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

FIG. 1 shows a hardware configuration of an electronic musical instrument according to an embodiment of the present invention. The electronic musical instrument shown in FIG.
A sound source (sound) is formed by using eight parameters consisting of four formant frequencies and four formant levels, using a formant synthesis type sound source of sound polyphonic pronunciation.

The voice (human voice) generated by a human is mainly formed of four characteristic formants. Therefore, if the sound is recognized simply as the pronunciation of "A", the four formants are converted to those. It suffices to give only the specified eight parameters to the sound source to make it sound. However, in practice, the formants are minutely fluctuating with time, and this alone cannot produce fine nuances of voice, and cannot express voice including consonants. Here, a plurality of eight parameters indicating the data of the instantaneous voice are prepared in time series in a storage circuit for each syllable, and are read out at each key-on. As a result, a voice such as "ga" can also be synthesized.

The judgment of multiple key presses is made as follows. 1. If there is a new key press from nothing being pressed,
The lyrics information read pointer forward by one, to reset the counter. 2. If the next key press is within the time counted by the counter, the pointer does not advance. 3. If a key press comes after the counter has finished counting,
This is regarded as a new key press, and the process 1 is performed.

The electronic musical instrument shown in FIG. 1 has a central processing unit (CP).
U) 11 is used to control the overall operation, and a read-only memory (ROM) 15, a RAM 17, and a sound source 31 are connected to the CPU 11 via a bidirectional bus line 13. The bus line 13 further includes a keyboard interface 19, an operation panel interface 23, and an operator interface 2 respectively.
The keyboard circuit 21, the operation panel switch group 25, and the operation element 29 are connected via 7. A speaker 35 is connected to the sound source 31 via a sound system 33 including a D / A converter and an amplifier (not shown).

The ROM 15 stores a control program for controlling the operation of the CPU and a syllable data sequence table. FIG. 2 shows a sequence table for syllable data and for lyric data described later. The sequence table for syllable data is a frequency sequencer (F-SE
Q) and a level sequencer (L-SEQ), each of which comprises four main formant frequencies F1 to F4 and a level (amplitude) which characterize each syllable (50 tones, muddy sounds, semi-dulled sounds, etc.) of each human voice. ) L1 to L4 are stored in chronological order.

The RAM 17 has a work area in which registers and flags for temporarily storing various data generated when the CPU 11 executes the control program are set, and a lyrics data sequence as shown in FIG. A sequence area in which a table (seq) is set is provided. The lyrics data sequence table (seq) can be constituted by various cards, tapes or disks, and a reader for reading data from them, other than the RAM.

The keyboard circuit 21 includes one or a plurality of key switches corresponding to each key (key) of the keyboard, detects the operation state of each key switch, and depresses each key.
Generates key information indicating key release and key touch (initial touch, after touch, etc.).

The operation panel switch group 25 includes various panel switches arranged on a panel surface of the electronic musical instrument, and generates switch information indicating an on / off state or a setting state of each switch on the panel surface. As a panel switch, a power switch, a tone selection switch, a volume setting switch, and the like are provided.

The operator 29 is composed of a pedal or the like, and generates operation information indicating each operation state.

The sound source 31 uses a formant synthesis type sound source disclosed in Japanese Unexamined Patent Application Publication No. Hei 2-262698, and is transmitted from the CPU 11 in response to inputs from the keyboard circuit 21, the operation panel switch group 25, and the operation element 29. Based on the generated musical tone formation information, musical tone (voice) waveform data for a maximum of 16 tones is formed by time division processing. This sound source 31
Is the amplitude modulation of the first repetitive waveform (modulated wave) by the second repetitive waveform (modulated wave), and the frequency of the first formant is mainly determined by the repetition frequency of the modulated wave, and the level and envelope of the The shape is mainly determined by the waveform of the modulated wave. The frequency of the second and subsequent formants and the relative level with respect to the first formant are mainly determined by the waveform of the modulated wave. Further, the pitch (pitch) of a musical tone is determined mainly by the repetition frequency of the modulated wave.

The sound system 33 performs D / A conversion of these musical sound waveform data and acoustically mixes them to generate a musical sound signal composed of a maximum of 16 sounds.
This tone signal is emitted as sound through the speaker 35.

Next, the operation of the electronic musical instrument shown in FIG. 1 will be described with reference to the flowcharts shown in FIGS.

Referring to FIG. 3, CPU 11 first executes initialization in step 301. In this initialization, a sequence step register step which is set in the RAM 17 and is used as a pointer for reading out a lyrics sequence, and a pointer step within the predetermined time in order to generate the same syllable by the key depression within a predetermined time after a new key depression. Clears registers and flags such as an elapsed time counter count used to prohibit stepping, a pedal flag pedal indicating the current state of the pedal, and TIME (ch) prepared for each musical tone forming channel. Initial settings such as a predetermined preset value are performed, and initial settings of the sound source 31 and other peripheral circuits are performed. Next, steps 303 to
Scan each interface at 319,
If there are events, the process proceeds to a routine for performing those processes. The sequence reading process of step 321 is as follows.
A set of formant data (4
This is a process for sequentially giving a sequence of two formant frequency data and four formant level data) to the sound source 31. The processing of steps 303 to 321 is repeatedly performed in a loop.

Hereinafter, the loop processing of steps 303 to 321 in the main processing routine will be described in more detail.

In step 303, a key scan is executed. That is, the keyboard interface 19 is accessed to check the operation state of the keys on the keyboard. Step 30
In 5, it is determined whether or not any key has changed (key event) based on the result of the key scan in step 303. If there is a key event, the sound generation process of FIG. 307 is executed (FIG. 4). If there is no key event, the process directly proceeds from step 305 to step 309.

At step 309, a panel scan is executed. That is, the operation state of the operation panel switch group 25 is inspected. In step 311, it is determined whether or not any operation panel switch has changed (panel event) based on the result of the panel scan in step 309. If there is a panel event, a panel process (not shown) in step 313 is executed. On the other hand, if there is no panel event, the process proceeds directly from step 311 to step 315.

At step 315, an operator scan is performed. That is, the operation state of the operator group 29 is inspected. In step 317, it is determined whether or not the operation state of any of the operators has changed (operator event) based on the result of the operator scanning in step 315. If there is a manipulator event, the manipulator process (FIG. 5) in step 319 is executed.
7 directly proceeds to step 321.

In step 321, after executing the sequence reading process (FIG. 8), the process returns to step 303.

FIG. 4 shows a flow chart of a sound generation processing routine. If it is determined in step 305 of the main processing routine that there is a key event, the CPU 11 executes this sound generation processing routine.

Referring to FIG. 4, CPU 11 first determines in step 401 whether the key event is a key-on event. If it is a key-on event, a read address setting process (FIG. 6) described later is executed in step 403 to obtain a read start address of the sequence memory. Thereafter, in step 405, a search for an empty channel is performed, and in step 407, the presence or absence of an empty channel is determined. If there is no free channel, a truncation process is performed in step 409 to forcibly form a free channel. In step 411, key-on information KON indicating that key-on is currently being performed is written to the empty channel ch searched in step 405 or forcibly formed in step 409. Thereafter, in step 413, the address set in the read address setting process in step 403 is written to the array register AD (ch) storing the read address of each channel. Further, at step 415, the last key register last-
The key code KC of the key that has been turned on is written in the key, and in step 417, a start pulse is sent to the envelope generator EG corresponding to the channel to start the output of EG. Then, in step 309 of the main processing routine, Reverse processing.

If the determination in step 401 is "NO", that is, if the key event in step 305 is a key-off event, it is first determined in step 421 whether a key-on key corresponding to this key-off key exists. This is because the tone may be forcibly dumped by the above-described truncation processing even if the player is on. Therefore, when the corresponding key-on key does not exist, the sound generation process is terminated and the process returns to step 309 of the main processing routine. On the other hand, if the corresponding key-on key exists, step 42
At 3, the key-off information KOFF is sent to the channel,
At step 425, key-on information KON of the channel
Is deleted, and the process returns to step 309 of the main processing routine after the sound generation processing is completed.

FIG. 5 shows a flowchart of the operator processing routine. If it is determined in step 317 of the main processing routine that there is an operator event, the CPU 11 executes this operator processing routine.

Referring to FIG. 5, in step 501, CPU 11 determines whether or not the operator event is a pedal-on event due to depression of a pedal. If it is not a pedal-on event, it is determined in step 503 whether or not the operator event is a pedal-off event caused by releasing the pedal. If the event is either a pedal-on event or a pedal-off event, the pedal flag pedal is inverted at step 505 each time the event occurs. Thereby, information on the current pedal is obtained in the pedal flag pedal. On the other hand, for an operator event that is neither a pedal-on event nor a pedal-off event, processing corresponding to the operator whose event occurred in the other operator processing in step 507 is performed. When the processing of step 505 or 507 is completed, the processing returns to step 321 of the main processing routine.

FIG. 6 shows a flowchart of a read address setting routine executed in step 403 of the tone generation processing routine of FIG.

Referring to FIG. 6, CPU 11 determines in step 6
At 01, the pedal flag pedal is checked. The pedal flag pedal is set so as to represent the current pedal information by the process of step 505 (FIG. 5).

If it is determined in step 601 that the pedal flag pedal is "1", that is, if the pedal is depressed, the syllable data sequence is executed in step 621.
The read address AD of the table (FIG. 2) is set to the same read address AD (l) as the last pressed key last-key.
After setting to (ast-key), the process returns to step 405 in FIG. In other words, the electronic musical instrument shown in FIG. 1 continuously emits the same syllable without advancing the lyrics even when a new key is pressed when the pedal is depressed.

On the other hand, if it is determined in step 601 that the pedal flag pedal is "0", that is, if the pedal is not depressed, it is determined in step 603 whether or not the count value of the elapsed time counter count is larger than "0". I do. Every time a new key is depressed, a value corresponding to the cycle of the interrupt counter, that is, a value corresponding to the tempo of the music, is set in the counter 611 described later, and the counter count is decremented by "1" each time an interrupt process described later is executed. (See step 705 in FIG. 7), and stores information on the lapse of time from the new key depression.

In the determination of step 603, the counter c
If the count value of “out” is “0”, the lyrics sequence read pointer “step” for reading the lyrics sequence table seq of FIG. 2 is incremented. That is, when the counter counts a predetermined time,
Advance the lyrics to the next syllable.

In the following step 607, the pointer step
Check the value of. If the value of the pointer step exceeds the last address seq-max of the lyrics sequence table seq, the pointer step is cleared in step 609. As a result, the pointer "step" indicates the head address of the lyrics sequence table seq, and the lyrics are read again from the beginning. For example, if each sound of "Halelujah" is stored in the sequence table seq one syllable at a time, the musical tone (lyrics) is repeated as "Halelujah" and "Halelujah" every time the key is pressed four times.

In the determination in step 607, the pointer s
The value of step is the last address se of the lyrics sequence table.
If it is equal to or less than q-max, the process of step 609 is skipped, and the process proceeds directly from step 607 to step 611. In step 611, as described above,
The counter count is set according to the interrupt interval time, which is the cycle of the interrupt counter, and then the process proceeds to step 613.

In the determination in step 603, when the count value of the counter count is larger than 0, it indicates that the time from a new key depression is within a predetermined time. In this case, steps 605 to 611 are skipped, and the process proceeds directly from step 603 to step 613. That is, in step 61, a new key is pressed.
The key pressed before the predetermined time stored in 1 is counted by the counter count is the same syllable as the new key pressed,
It is used as musical tone formation data of different pitches, and a double tone performance by voice is realized.

In step 613, the read address AD
Is the pointer step of the lyrics sequence table seq
Is set to the data seq (step) at the position indicated by, that is, the head address of the syllable data indicating one sound of the lyrics, and the process returns to step 405 in FIG.

FIG. 7 shows the counter c used in FIG.
9 shows an interrupt processing routine for automatically incrementing “out”.

Referring to FIG. 7, CPU 11 is supplied with an interrupt clock of a predetermined cycle from a timer (not shown), and executes this interrupt processing routine each time this clock is supplied. Here, in order to prevent multiplexing of interrupts, this interrupt processing routine is started (step 70).
At 1) and at the end (step 707), interruption is prohibited and permitted. In step 703, the count value of the counter "count" is checked. If the count value "count" is larger than 0, the count value "count" is decremented in the next step 705. In step 703, the count value co
If it is determined that unt has already become 0, step 705 is skipped and the count value count is kept at 0.

FIG. 8 shows details of the sequence reading routine executed in step 321 of FIG.

Referring to FIG. 8, since the electronic musical instrument of FIG. 1 has 16 musical tone forming channels, CPU 11 first sets "16" in general-purpose register i in step 801. In a succeeding step 803, it is determined whether or not the key of the i-th channel is on. If the i-th channel is key-on, the value ctr of the running counter is checked in step 805. Running counter value c
If tr is equal to or greater than TIME (i), which is time information since the previous change of the formant sequence, the read pointer AD (i) of the syllable sequence is advanced, but the formant sequence assigned to each syllable is advanced. If the number exceeds the number of sets, the formant information of the syllable arranged next in the syllable sequence is read irrespective of the lyrics, which is inconvenient. Therefore, in step 805, the pointer AD (i) is checked, and if the value has reached the maximum value MAX of the pointer AD (i), the processing of step 807 is skipped to thereby determine the pointer AD (i).
The progress of (i) is stopped. On the other hand, if the value of the pointer AD (i) is smaller than the maximum value MAX, the pointer AD (i) is incremented in step 807. In step 805, the end point of the data is managed by the number of data. However, the end point may be determined using special data such as an end mark. In this case, the data size for each syllable can be made variable.

When the sequence read address AD (i) is determined, in the next steps 811 and 813, based on the address value AD (i), the formant frequencies F1 to F4 and the fault levels L1 to L are obtained from the table.
4 are read out and sent to the sound source 31 in step 815. Further, after updating the elapsed time information TIME (i) according to the speed of the running counter in step 817, the process proceeds to step 819.

If the result of determination in step 803 is that the i-th channel is not key-on, there is no need to process the sequence information for that channel, so that steps 805 to 817 are skipped and step 803 is directly executed. The process proceeds to step 819.

Also, as a result of the determination in step 805,
TIME (i) in which the value ctr of the running counter is time information since the previous change of the formant sequence
, The sequence information for the i-th channel does not need to be changed yet, so that the processing of steps 807 to 817 is skipped and step 80 is skipped.
5 directly proceeds to step 819.

At step 819, the general-purpose register i is decremented. At step 821, it is determined whether or not the value i of this register has become "0". If the value i of the register is "0", this means that the sequence reading process has been completed for all 16 channels, and the process returns to step 303 in FIG. On the other hand, if the value i is not "0", there are channels for which the sequence reading process has not been completed yet, so the process returns to step 803, and the processes of steps 803 to 819 are repeated. .

In the above embodiment, when the sequence step register step reads out the lyrics to the end,
Although it is set to be cleared, this register step
May be forcibly cleared. As a result, it is possible to clear the desired state, for example, when the step has been advanced due to an erroneous performance.

When the duration of key-on is less than a certain level, the step may not be advanced so that slight mistouch or chattering may cause erroneous performance.

A plurality of sequences seq may be switched using another operator.

A different sequence may be read out by applying the technique of key range division. In this case, a performance such as a round singing can be realized.

FIG. 9 shows a hardware configuration of an electronic musical instrument according to another embodiment of the present invention. In this figure, parts common or corresponding to those in FIG. 1 are denoted by the same reference numerals.

The electronic musical instrument shown in FIG. 9 has a function of simulating a desired sound, and is a formant comprising a RAM having a formant sequence table area of the same format as the formant sequence table set in the ROM 15 of FIG. It includes a sequencer 37, an analyzer 39 for comparing a musical sound formed by the sound source 31 with a reference sound, and a display 41.

The analyzer 39 performs fast Fourier transform on PCM data such as voice input from a microphone and formant waveform output from the sound source 31 to create their spectrum information, that is, data on formants (formant information).

The display 41 displays the spectrum of the voice and the musical sound or the difference between these spectra.

The formant sequencer 37 records data relating to the formant obtained by analyzing the voice, for example, the center frequency (formant frequency) of the formant, the level, the ratio U / V of the vocal sound U to the unvoiced sound V, etc. as time-series data. For example, data pronounced as "Thai" is recorded in the formant sequencer 37 as shown in FIG.

In FIG. 10, (a) shows the formant frequencies F1 to F4, and (b) shows the formant levels L1 to L4.
, (C) is the ratio U / V between voiced and unvoiced sounds.

In the embodiment shown in FIG. 1, data is read from the point at time 0 at the same time as the note-on, and the sound is generated. In this embodiment, the start point read at the same time as the note-on is changed by the velocity. I have.

For example, to convert velocity data (0 to 127) into a read point, a sensitivity parameter sense is provided, and read-point = v
An operation of “elocity-data × offset” is performed. Here, when the keystroke is stronger, the readout point is increased, sense ≧ 0, and when the keystroke is stronger, the readout point is decreased, sense <0.
Set to 0. However, in the latter case, since the read-point becomes negative, it is necessary to set the offset.

In FIG. 9, the formant sequence table in the ROM 15 is read by the formant sequencer 37 when used, as a table set by the manufacturer.

In order to change the read point of the formant sequencer 37 by the key velocity,
For example, in the operation of the electronic musical instrument of FIG.
13 may be changed as in step 413 'of FIG. That is, in step 413 ′ of FIG.
The value obtained by multiplying the key velocity velocity by a predetermined sensitivity parameter sense is used in step 6 in FIG.
13 is added to the read address set as the offset, and the tone formation information read counter AD
(Ch) is written as a read address read-point.

As described above, in the electronic musical instrument shown in FIG. 9, by changing the readout point of the formant sequencer 37 according to the key velocity, the timbre can be dynamically changed according to the velocity.

For example, assuming that the analysis data which pronounces “Thai” is recorded in the formant sequencer 37,
If the read point is changed according to the velocity, the sound is pronounced "tie" when played strongly, but when played weakly, the information of the consonant part is skipped and pronounced "eye".

This can also be effectively applied to the case where formant analysis data of musical instrument sounds is recorded in a formant sequencer. For example, creating a timbre change without manipulating timbre parameters, such as reading the part where the formant of the rising part of the wind instrument changes drastically with velocity, such as `` buo '' when playing strongly and `` poor '' when playing weakly Can be.

The electronic musical instrument shown in FIG. 9 can be developed as follows.

For example, as shown in FIG. 12, a plurality of musical tones ("Wah", "Har",...) Correspond to each velocity range (0 to 24, 25 to 48,?, 100 to 127).
"Woo") is recorded in the formant sequencer 37, and the read start address rea
The d-point may be changed so that the memory number is selected by the velocity data.

That is, when the key is hit at the velocity 45, the memory number Mem-num3 is selected as the read start address read-point, and the formant sequencer 37 reads out the formant information for generating "NAR".

By doing so, completely different sounds can be produced with the same timbre parameters by the velocity data. In the above case, by gradually increasing the velocity, the pronunciation can be changed in the order of "Wah" → "Har" → "Nah" → "Ah" → "Wh".

That is, when performing a backing chorus or the like using an electronic musical instrument (formant sound source), the sounding pattern can be changed by the key velocity.

In the above description, operator information of another operator such as a modulation wheel or a pedal may be used instead of the key velocity.

Another embodiment of the present invention is as follows.

Further, there is provided performance operation information input means for inputting performance operation information other than the pitch designation information, and a read point when the reading means reads the formant information from the formant information storage means is set to the performance operation information. 2. The electronic musical instrument according to claim 1, wherein the electronic musical instrument is changed according to the following.

[Brief description of the drawings]

FIG. 1 is a block circuit diagram illustrating a hardware configuration of an electronic musical instrument according to an embodiment of the present invention.

FIG. 2 is a schematic format diagram of each sequence table in the electronic musical instrument of FIG.

FIG. 3 is a flowchart showing a main processing routine in the electronic musical instrument of FIG. 1;

FIG. 4 is a flowchart showing a sound processing routine in the electronic musical instrument of FIG. 1;

FIG. 5 is a flowchart showing a control processing routine in the electronic musical instrument of FIG. 1;

FIG. 6 is a flowchart showing a read address setting processing routine in the electronic musical instrument of FIG. 1;

FIG. 7 is a flowchart showing an interrupt processing routine in the electronic musical instrument of FIG. 1;

FIG. 8 is a flowchart showing a sequence read processing routine in the electronic musical instrument of FIG. 1;

FIG. 9 is a block circuit diagram showing a hardware configuration of an electronic musical instrument according to another embodiment of the present invention.

FIG. 10 is a graph showing an example of time-series data stored in a formant sequencer 37.

FIG. 11 is a flowchart showing a sound processing routine in the electronic musical instrument of FIG. 9;

12 is a diagram showing formant analysis data in a formant sequencer in a modification of the electronic musical instrument of FIG.

[Explanation of symbols]

11: Central processing unit (CPU), 15: ROM, 17:
RAM, 21: keyboard, 31: sound source.

──────────────────────────────────────────────────続 き Continuation of front page (51) Int.Cl. ⁶ identification code FI G10L 9/04 G10L 9/04 G (58) Investigated field (Int.Cl. ⁶ , DB name) G10H 1/24 G10H 1 / 053 G10H 1/18

Claims (1)

(57) [Claims] And pitch designation information generation means for generating 1. A pitch designation information, the formant information storage means for storing each of the time-series information of a plurality of formants characterizing calling sound of the human vocal, before Kion high specified Reading means for sequentially reading the time-series information in a predetermined sequence in response to generation of the pitch designation information from the information generating means , wherein the pitch designation information is
If it occurs continuously within the time of
Do not proceed with the above sequence except for the pitch designation information
And casting, every each pitch designation information generated from said pitch designation information generation means, formant forming source for forming a sound based on the sequential formant information is read by the sound high specification information and said reading means An electronic musical instrument comprising: a plurality of sounds that can be formed in parallel.