JPS5837693A - Electronic melody instrument - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] This invention provides m! ＋、'lπ1・～(This work relates to an electronic singing instrument that can be used to sing with a human voice.

In recent years, the keyboard has become heavier and heavier, like an electronic organ.

Attempts have been made to make the song sing using the Hitoni i sound.

For example, a specially provided typewriter-style phoneme designation switch, or electronic music :): Rinoh key switch can be distributed to each note 11 ↑ 1, and the phoneme f) l constant switch, etc. Specify the lyrics sequentially in advance and store them in the device (RA, M), and in the performance mode, record the lyrics in accordance with the key press timing. Read sequentially; 1”T depending on key information
]i'i+ (pitch) to generate a singing voice f
An electric musical instrument has been proposed (see Japanese Unexamined Patent Publication No. 77799/1983).

However, such electronic instruments require a large number of ! to memorize the lyrics. 'T: There was a problem in that it was necessary and confusing to operate the rli determination switch, and the operation was cumbersome.

This invention was made to eliminate such a pause: 11fJ by Iq'C. Before the performance, the lyrics are manually written in one voice, and the phonological data is written down. 1. Go to performance mode, 1.I)l 1. (formerly, 'I) output the phonological data and specify the pitch.
Lyrics written in advance by voice ``ノ〕 Boat ni people...: i j'7
It is a singing electric device that can make people sing a song by making them sing.

Hereinafter, the contents of this invention will be explained using a real example of 1/11.

The 11th is this? It is an inkstone drawing (example 7 r1.
In the -L part of the microphone held by the frame tube 2,
A keyboard 4 is provided on the front side of the main body 1 and has a large number of key switches corresponding to pitches.
Bay Le Part 1a (OK, Saturday, Speaker J + 7j1.,
'+ (, ':\, Switching mode 9 mode SW/S F'9
Outside (・1・1rll up switch 6, note: d, key of tone, etc.': t'tllll Voluno・7, (ke''j!'
Next to I□i4
It is.

FIG. 2 is a block diagram showing an example of the interior of the main body 1 at the 11th point (t6).

This communication processing system manually detects the Mi voice signal from the microphone 3 as the first stage of voice input, and identifies the 1st attack 11).
1" quality as a means of phoneme discrimination: 11 infatuation" IQ equipment 1
0 and the phoneme data identified by it:'
I'm old. 111・i,・:”i, It as T stage,
AJV 1i a(r., self・[,・two,'N
lj311 and the closing Ji which is opened and closed by operating the keyboard 4.
The key data generation circuit 16 which outputs the 1゛'''] quotient data DI) and the data read signal -;' 2. Sound Δ that outputs a sound source signal with a corresponding pitch, Book 1-11, 14 and 1. Request・1・:\
i''l is sequentially determined by the 1"71 yen data output from the circuit 11 and the sound source count output from the sound source circuit 14.
Speech synthesis device 15 that synthesizes and outputs '+'・1,1
Then, the outputted voice signal is amplified by an amplifier (-), and the output signal is amplified by an amplifier.

It consists of a sound instrument 16 that adjusts the tone color and the like and outputs the sound from the speaker 5 of the 11th speaker 1 (knee).

In addition, SWl is a storage mode switch, and 8W2 is a steam mode mode switch. In addition, the double-crossed line 9 in the figure indicates a small number of coded digital marks.

An example of this (second, first write down the lyrics you want to sing)
3-, l, , ::41, when doing so, follow the notes 1. i::Close i-mode switches S to IVI and proceed as described in 1. C"(L circuit 1
1 to enable data writing.

Then, if you say the lyrics into the micro,
□J: i is converted into an electrical signal by the micro
-') Voice 1j'l;! A-/T
) The converted suffix ρ number spectrum is converted into 1 loaf, the standard phonetic pattern and o > il conversion 4, q are performed to form “A-1 “I”.
"tsu"... "n" etc. 1'''1'1''iL1
is identified, and from the fame recognition device 1o, the one phoneme (d) ``stomach'' 111 data l)s and the data input/output signal s are 1
It is neutralized.

Accordingly, the storage circuit 11 receives the data intrusion signal 'I'.
'Every time the JSW is manually input, the phonological data I)s is
, AM lia's address, 1st child's address (2nd place 1
Missing, 7self...; Makes you think.

Next, 71. ! When playing i, 4jτ playing mode switch +S
Close W2 and write, l,? The i-circuit 11 is enabled for data reading.

And 7. :i <%-J- If you turn the keyboard 4 to 1.lv::('l) according to the melody of the song and open and close the meter switch 12, the data read signal SR will be generated from the key data generation circuit 16. Sequential reading of phoneme data from the memory circuit 11 = 4- and pitch data l)p? live sound i11;j
From the circuit 1/1, the key switch 12 (2 by 1', i'+') island 1'44 source signal is output.

In addition, the sound synthesis device 15 sequentially synthesizes manually generated sound source signals into (') lines/1ij based on phoneme data.
( ): 1i''I- is switched, octagonal formation is performed by a filter according to the vocal tract model, etc., and the voice signal is formed into i''7 and output.

Then, the sound system 16 pronounces the greeting.

In this way, the lyrics input in advance by voice) can be
Keyboard operation C21', 1 degree specified with the key switch
:You can sing a song by using the human voice of i.

FIG. 6 is a block diagram C showing another embodiment of the 1.1-inch processing system for an electronic singing musical instrument according to the present invention, and the same parts as in FIG. [It's shi.

In this example, the first example is l', 'l'j"'F
Record the device phonological data as is from the percept,
,7j, so if the song, ,lll is long, the human world [1
This was done in consideration of the need for 1 memory, and the manually input sounds j', ', -j (75j's ('''1. H
Instead of storing only ``111111'' data (2. By memorizing the phonetic verbs set in advance for each ``111'' and made with i, 0)'N:1: can be small. This is how it was done.

sl ``wa l:), each phoneme with the 1'' position odd, e.g.
・・Determine ``Iga 1' fi 'J-amba for N1,''
Based on this phonetic number and 71, the audio data of h1゛'portion'' loss (frequency s''1゛shi1.', q) and height l;1 of the data storage circuit 20 are k. 0 to 42021 and leave I 曾已'lC\sa1.

71・T, pitch,..., H, voice data from H recognition device 10'I) A and 1'''lI:・I-Riuraura・:1・,Audio data read from time i1'820 1)!(and press (l し(7, when both are the same or bu, n similar, f knee 1・(
I1、゛bow 1・:(ﾾI1 makes the sound J'・J-ru sound I:I
Data comparison circuit 21, match 1. ．． ! J'l.: (Sequentially record the phonetic numbers to 11 and 8 when manually written in 122a. 1.: χ, and set the data generation circuit 10 or 1') Carta readout Ij1' S R is manually written every l+,
AM 22: Recorded on +1. ,,:l; Read out the phoneme number that is being used l″ll′:if number record・1.・:IL
A circuit 22 is provided.

Zunawa f2), in this real example, the voice recognition device 1
0' and 1°'-1 data comparison circuit 21 constitute a phoneme identification means iI"i, and the phoneme 1:1 data storage circuit 20
The 1m number memory circuit 22 records and registers the 1m number memory circuit 22.

In this example, write down the lyrics you want to sing: l', t
To do this, close the memory mode switch SWI and play the lyrics into the micro.

Then, the micro-)te mountain: <Cf, j"J
The voice 1 recognition circuit 10' is set to high 7.
:ji(,,Voice data 1 that was manually analyzed)
Δ becomes 1'1° mountain deg'lj'j
1111 data storage circuit 20 at high speed! 'if
'J' number No. corresponding sound, lj, lj data [) 13 is read out 11.1 Phonological number No. Har'I
-1''iii ;'-Nha'j, ilr times f'32
'l +=, i"1'r'F Data T) B is sound T
Knee i data 1 punch 1 ill road 21 friend noan moon.

The audio data comparison circuit 21 performs the following:
Compare A and I) B, and both B are the same - or 1, ri similar 1
Match Ire';'j EQ is output to 111. 1), the circuit 22 to the phonetic number register 'I' then converts the Δ applied l'explode 11 number NO to ILA-M2.
In 2a, write i,,,':de.

In this way, the sound of 7′l:, j+, ′
jli j-ter 7- in sound i:Ii nounba ++1t1st order R, A, M 22
a (: remember.

To 11, If, close performance mode switch SW2 and operate l'11'A'k 4, key switch 1
2 opening/closing [2] By the data read signal No. 5'Sr+ generated from the GEE data generation circuit 13, the phoneme number storage circuit 22 reads If, AM 22a, and the stored sounds f'', ii: J-y. The sound j)-1 is read out sequentially.
The data storage circuit 2o is manually inputted, and the audio data corresponding to the sound J:j is read out to the sound J:j synthesizer 15.
to use human power.

The operations of the 8-source circuit 14, 1''♀l)f synthesizer 15 and sound system 160 are the same as those of the embodiment shown in FIG. 2, so a description thereof will be omitted.

In this fruit 1f [by time nl -], similarly to the previous embodiment, the lyrics are manually memorized using the )'7 mountain, and the keyboard operation (by specifying i゛1-j:':+ from 2) is performed. You can generate a human voice and have it sing a song on any given day.

t"Oh, for example, the speech recognition device and convex, jj synthesizer in these real examples are based on the partial autocorrelation coefficient (1
'A Il, COII, coefficients) into 5J (, r,
PC) type vocoder 8-, etc. can be used.

As explained in Theory 11J above, this singing pattern 1' based on Demei ζ2
The musical instruments are Pme1゛♀, (I wrote the lyrics by I7, 1.
, j:i, and when performing, by Ken) 14-inch operation, [t':':'s melodrama ``Person's throat according to the lyrics memorized in advance on L day''? g) can be made 1, so the same song, 1Iil, can be made i:! in the middle of the chorus; stomach and ii)'Ki
When used for composing or arranging music for J1, Q-ii2 is very convenient, and the operation is +mrii-.

[Brief explanation of the drawing]

FIG. 1 is a perspective view showing the external appearance of a medium-sized 1-inch singing electric instrument according to the present invention, and FIG.
FIG.
This is a block diagram similar to 1"<1. 1...Main body 610..-, Iku4...
Keyboard 5...Speaker 10, 10'...
Voice recognition device 11...Memory circuit 12...Misaki switch J~16...
...Gee data generation circuit 15...Convex, ll/↑] Device 20...1゛° box: I data record, 1. ．． ．． xll,
Circuit 21. . Takano (l data ratio 1 bite circuit

Claims (1)

[Claims] 1 The human-powered voice means to manually make the song 1i#I with l"i'J'i i, and this one. a storage means for storing the 1'' day 111 data identified by the phoneme identification means; a keyboard having a large number of key switches corresponding to -;rl, , HJ, ti;
7''i'', which sequentially reads out the phoneme data stored in the storage device when the ``)'' was created, and outputs the pitch pitch specified by the key switch; and 7''i'';
, i'5 signal-; a generating means and a speech synthesizer that sequentially synthesizes and outputs 111 inches of speech using the 1"1j1 data read by the means and the generated pitch signal. Eh, [stock] 1! ', lK 1.Y work ('''
- An electronic singing instrument that allows you to sing a song with lyrics that have been manually written by voice.