JPH04166999A - Acoustic signal synthesizing circuit - Google Patents

Abstract

PURPOSE:To cause sounds to be generated before and after synthesis in a smooth successive manner and also enable sounds approximating natural conditions to be synthesized by providing a plurality of envelope circuits each of which generates a sound in synchronization with the timing of generation of acoustic data, a D/A converter for D/A converting acoustic data from a storage circuit, a selector and a mixing circuit. CONSTITUTION:Envelope circuits 4, 5, 6 each generate an envelope signal corresponding to acoustic data about a first sound simultaneously with generation of this acoustic data and generate another envelope signal corresponding to acoustic data about a second sound simultaneously with generation of the acoustic data so as to perform envelope processing and at the same time the first sound damped by the envelope signal at that time is synthesized. When a third sound is generated an envelope signal corresponding to acoustic data about the third sound is generated likewise and the third sound is synthesized with the first and second sounds damped by the respective envelope signals at that time. An envelope signal corresponding to the acoustic data of each sound is thus generated and as long as the sound continues a sound damped by the envelope signal is synthesized so as to obtain a synthetic sound in approximating natural acoustic conditions.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an audio signal synthesis circuit that synthesizes sounds such as melodies.

[Prior Art] As an acoustic signal generator of this type, for example, Japanese Patent Laid-Open No. 59-26789 has been known. Normally, when a single sound is uttered, the timbre of the same sound changes slightly over time. That is, the high frequency components included in a sound are easily attenuated, and the low frequency components are difficult to attenuate, so the state (timbre) of the sound changes slightly over time. Therefore, in the above-mentioned conventional system, the waveform of the tone when the sound is emitted, the waveform of the tone after it has decayed after a predetermined period of time has passed since the sound was emitted, and the waveform of the tone after the sound has attenuated after a predetermined time has elapsed, and then Three types of acoustic data, including the waveform of the tone after attenuation, are set in advance and stored in each of the three memory circuits (ROM), and the three tones that are emitted sequentially are created using these three types of acoustic data. The three sounds are synthesized by modifying the waveforms, digitally summing the outputs, converting the outputs from D/A, and applying one envelope signal to them in common.

[Problem to be solved] In the configuration in which sounds are synthesized in this way, for example, as shown in FIG. However, if we consider the case where the third sound A is further overlapped at timing T3 to create a melody, when the second sound B is emitted, the first sound
Even though the sound C is attenuated to have an amplitude of C1, the amplitude changes to C2 when the second sound B is generated.

This is due to the fact that a common envelope signal is added to sounds B and C, and when the second sound B is generated, envelope processing is performed so that it has the maximum amplitude.
The amplitude of sound C should be 01, but it increases to 02. In other words, regardless of the length of the emitted sound, it is always treated as being in a predetermined attenuation state, so it is rare for the amplitudes c1 and 02 to match, and in the example shown in Figure 5, the amplitude increases from C1 to 02. As a result, the sound becomes discontinuous and lacks smoothness for the listener. Similarly, when emitting the third sound A, sound A is further added to the attenuation state of sound B and sound C, but even at this time, the predetermined attenuation state is always maintained regardless of the length or shortness of the emitted sound. Since the third sound A is added to the sounds B and C, the amplitude b1 and the amplitude C at which the second sound B is actually attenuated are
It is rare for the amplitude c3 where 2 is actually attenuated and the added amplitude b2°C4 to match, and in the example shown in Figure 5, it decreases discontinuously from bl to b2, and discontinuously from C3 to 04. The noise increases and adds up, making the sound sound less smooth.

Also, after multiple sounds are synthesized, 1
Since processing is performed using two envelope signals, the ratio of the loudness of each sound is always the same, and the ratio of the loudness of each sound at the same point in time becomes constant. This is quite different from the natural acoustic state.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to enable the synthesis of sounds in which the sounds before and after synthesis are smoothly continuous when a plurality of sounds are synthesized, and which is close to the natural state.

[Means for Solving the Problems] In order to achieve the above object 1-1, the audio signal synthesis circuit of the present invention includes a plurality of storage circuits each storing different audio data, and a plurality of storage circuits each storing different audio data, and A plurality of envelope circuits generate corresponding envelope signals in accordance with the generation timing of each of the above-mentioned acoustic data, and D/A converts the acoustic data from each of the above-mentioned storage circuits, and corresponds to the corresponding acoustic data from each of the above-mentioned envelope circuits. A plurality of D/A converters perform envelope processing using the envelope signal (=), and the envelope signal from each of the envelope circuits is sent to the D/A converter to which the audio data is supplied.
It includes a selector for outputting to the A converter, and a mixing circuit for synthesizing the output signals from each of the D/A converters.

The selector selects the acoustic data from each of the storage circuits,
The arrangement may be such that the envelope signal corresponding to the acoustic data is output to a D/A converter to which it is supplied.

[Operation] The envelope circuit generates an envelope signal corresponding to the acoustic data of the first sound at the same time as the acoustic data of the first sound is generated, and generates an envelope signal corresponding to the acoustic data at the same time as the acoustic data of the second sound is generated. At the same time, the sound attenuated by the envelope signal of the first sound at that time is synthesized. Also, when the third sound is generated, an envelope signal corresponding to the acoustic data is simultaneously generated in the same manner as above, and the sound is attenuated by the respective envelope signals of the first and second sounds at that time. Synthesize. In this way, by generating an envelope signal corresponding to each sound data and synthesizing the sound that is attenuated by the envelope signal while the sound continues, a synthesized sound that is close to a natural reverberation state is obtained.

[Example] An example of the present invention will be described based on FIGS. 1 and 2.

First, as shown in the first diagram, a memory circuit (ROM) 1.2.
3 stores different acoustic data (for example, PCM data), for example, acoustic data at the beginning of a sound, acoustic data after a short time from the start of the sound, and acoustic data after a further period of time has elapsed. Each data is memorized. The number of memory circuits is matched to the number of tones to be synthesized, and in this example corresponds to three tones. A plurality of envelope circuits (CR) 4, 5.6 are provided in order to generate envelope signals corresponding to the audio data in accordance with the generation timing of each audio data.

The T terminal, which is one input terminal of each of CR4, CR5, and CR6, receives timing signals Tl, T2. T3 is input, and the other input terminal, A terminal, receives a start signal al at each timing from the control circuit 7.

a2. a3 is supplied. Output terminal Q of CR4, 5, 6
From start signals al, a2. At the same time as a3 human power, timing signals TI, T2. Envelope signals ql, q2゜q3 having different waveforms are outputted by T3. CR
Envelope signals Ql from 4, 5, 6, q2.
q3 are input terminals DI, D2 . are supplied to D3 respectively.

Input terminals Di, D2 . D3 and output terminals Ql, Q2. The connection with Q3 will be explained later in detail, using the timing signal T supplied from the control circuit 7.
l, T2. It is switched in response to T3.

A memory circuit (ROM) 9 stores frequency division ratio data for specifying the pitch (scale) of each note of a melody to be played, and is controlled by the control circuit 7 to specify the division of a specific note. The frequency ratio data is output and latched by one of the latch circuits 10, 11, and 12 which receives a latch signal from the control circuit 7. The number of latch circuits is equal to the number of sounds to be synthesized (three in this example). A sufficiently high fundamental frequency signal F is supplied to the variable frequency divider circuits 13, 14.15 provided corresponding to each latch circuit, and this frequency signal F is supplied from the latch circuits 10, 11.12. A clock signal for obtaining a desired musical scale is obtained by dividing the frequency based on the frequency division ratio data. [Variable frequency dividing circuit] 3.14.15 is supplied to the input terminals DI', D2°, and D3' of the second selector 16. Input terminal D1° in second selector] 6,
D2°, D3゛ and output terminals Ql', Q2°, Q3°
As will be explained in detail later, the connections between the timing signals TI, T2 . It is switched in response to T3.

Output terminals Ql', Q2°, Q3 of the second selector 16
The outputs of .degree. are counted by counters 17.18 and 1.9 and address the memory circuits (ROM) 1 and 2.3. The outputs from ROMI and 2.3 are each D/A.
The converters 20, 21, 22 perform D/A conversion, but at the same time, the output terminals Ql, Q2, . Envelope processing is performed using the envelope signal from Q3.

The outputs of the D/A converters 20, 21, and 22 are combined in a mixing circuit 23, and sent to a speaker 25 via an amplifier 24.
The sound is emitted from.

Next, the operation of sound synthesis will be explained with reference to FIG.

Here, the sound C emitted at timing signal T1 is a quarter note "do", then the sound B emitted at timing T2 is a half note "shi", and then the sound A emitted at timing T3 is an 8 This is the diacritic note "mi", and the case where these are sequentially synthesized to form the melody of "do-re-mi" will be explained.

In order to emit the first sound C, from the control circuit 7 CR4,5
,6,ROM9. Timing signal T to latch circuit 10
1 is supplied. Therefore, frequency division ratio data for specifying the scale of the note "do" is output from the ROM 9, and the latch circuit 1
The frequency signal F is latched at 0, and the frequency signal F is divided by the variable frequency dividing circuit 13 to produce a clock signal for obtaining the "C" scale, and is supplied to the terminal D of the second selector 16.

When the second selector 16 receives the timing signal TI, Ql'=
Since they are connected to DI', Q2'=D3', and Q3'=D2', the above clock signal is supplied to the counter 17, and the acoustic data of the beginning of the sound of "do" is read out from the ROMI by its output.

On the other hand, at the same time that the timing signal T1 is supplied from the control circuit 7 to CR4, 5, and 6, the start signal a1 is supplied to CR4.
Since the envelope signal q1 of the timing signal TI is generated from CR4, it is supplied to the terminal DI of the selector 8. The selector 8 is activated by the timing signal T1.
Q1 = Dl, 'Q2 = D3, Q3 = D2, so the acoustic data of the beginning of the sound of "do" from ROM1 is D/A converted by the D/A converter 20, and at the same time, the sound data of the selector 8 is Envelope processing is performed using the envelope signal q1 of the timing signal TI from the terminal Ql.

The signal from the D/A converter 20 is sent to the mixing circuit 23
．． Sound is emitted from the speaker 25 via the amplifier 24.

11 Next, in order to emit the second sound B, from the control circuit 7
R4,5,6,ROM9. Timing T2 is supplied to the latch circuit 11. As a result, frequency division ratio data for specifying the scale of the sound "shi" is output from the ROM 9, which is latched by the latch circuit 1], and the frequency signal F is divided by the variable frequency divider circuit 14 to produce the sound "shi". A clock signal for obtaining the musical scale is created, and this is sent to the terminal D of the second selector 16.
2°. The second selector 16 receives the timing signal T2 so that Ql'-D2', Q2''=Dl', Q
3"=D3°, the above clock signal is output from the input terminal D2° to the output terminal Q1°, and is further supplied to the counter 17, and the acoustic data of the beginning of the sound of "shi" is read out from the ROMI. , the clock signal for obtaining the scale of the first note "C" from the variable frequency divider circuit 13 is outputted from the input terminal D1° to the output terminal Q2', further supplied to the counter 18, and the clock signal for obtaining the scale of the first note "C" is output from the ROM 2. Acoustic data a little after the beginning of the sound is read out.

On the other hand, at the same time that the timing signal T2 is supplied from the control circuit 7 to CR4, 5, and 6, the start signal a2 is supplied to CR5.
Since the envelope signal q2 at timing T2 is generated from CR5, it is supplied to the terminal D2 of the selector 8. The selector 8 selects Ql according to the timing signal T2.
=D2, Q2-DI, Q3 =D3 The acoustic data of the beginning of the second sound "shi" from Renote and ROMI connected to D/A Ko>/(-D at 20 pm)
/A conversion, the timing from the output terminal Ql of the selector 8 (envelope signal q2 of No. j T2
Envelope processing is performed by Also, the acoustic data a little after the beginning of the first sound "do" in ROM2 is D.
At the same time as being D/A converted by the /A converter 21, envelope processing is performed by the envelope signal ql of the timing signal T2 from the output terminal Q2 of the selector 8. D
The signals from the /A converters 2o and 21 are synthesized in a mixing circuit 23, and the sound "shi" that started sounding and the sound "do" that started sounding a little earlier are synthesized and output from the speaker 25 via the amplifier 24. left alone.

13 Next, in order to emit the third sound A, the control circuit 7 to C
R4,5,6,ROM9. A timing signal T3 is supplied to the latch circuit 12. As a result, frequency division ratio data for specifying the scale of the note "Mi" is output from the ROM 9, which is latched by the latch circuit 12, and the frequency signal F is divided by the variable frequency divider circuit 15 to produce the scale of the note "Mi". A clock signal is generated to obtain the terminal D3 of the second selector 16.
'Supplied to. The second selector 16 receives the timing signal T.
3, Qlo−D3°, Q2°=D2°, Q
3°=D1°, the clock signal for obtaining the third note "mi" scale is output from the input terminal D3' to the output terminal Q1', and is further sent to the ROM via the counter 17.
At the same time as the acoustic data of the beginning of the sound of "mi" is read from I, the clock signal for obtaining the scale of the second sound "shi" is sent from the input terminal D2' of the second selector 16 to the output terminal Q2.
The acoustic data a little after the beginning of the sound of "shi" is read out from R, 0M2 via the counter ]8, and the clock signal for obtaining the scale of the first sound "do" is It is output from the input terminal D1゛ of the second selector 16 to the output terminal Q3゛, and further via the counter 19 to the RO
From M3, acoustic data a long time after the beginning of the C sound is read out.

On the other hand, at the same time that the timing signal T3 is supplied from the control circuit 7 to CR4, 5, and 6, the start signal a3 is supplied to CR6.
Since the envelope signal q3 of the timing signal T3 is generated from CR6, it is supplied to the terminal D3 of the selector 8. The selector 8 is activated by the timing signal T3.
Since QI=D3゜Q2=D2 and Q3=DI are connected, the acoustic data of the beginning of the third sound "mi" from ROM1 is D/A converted by the D/A converter 20, and at the same time, Envelope processing is performed using the envelope signal q3 of the timing signal T3 from the output terminal Q1 of the selector 8. Furthermore, the waveform of the acoustic data from the ROM 2 a little after the start of the second sound "shi" is from the D/A converter 2.
1, the envelope signal q2 of the timing signal T3 from the output terminal Q2 of the selector 8 is
Enveloped by Further, the waveform of the acoustic data from the ROMB after a long time has elapsed from the start of the sound of the first sound "do" is D/A converted by the D/A converter 22, and at the same time, the timing signal T3 from the output terminal Q3 of the selector 8 is converted.
Envelope processing is performed using the envelope signal q1 of . The signals from the D/A converters 20, 21 and 22 are combined in the mixing circuit 23, and the sound "mi" which started sounding, the sound "shi" which started sounding a little before this, and the sound "shi" which started sounding a little before this, are synthesized by the mixing circuit 23. The sound "do" is synthesized and emitted from the speaker 25 via the amplifier 24.

In this way, in the configuration shown in the first diagram, each sound continues to be emitted while being attenuated by the envelope signal that is generated in accordance with its own generation timing from the beginning of the sound until the third sound is generated. It does not increase or decrease discontinuously. The envelope signal is
Since three types are created, the number of which is equal to the number of sounds to be synthesized, and which match the generation timing of the sounds to be synthesized, it is possible to obtain a natural synthesized sound.

= 16- In the second embodiment shown in FIG. 3, the second selector 16 of the first embodiment is omitted, and in the first embodiment, latch circuits 1.0, 11, and 1.2 are provided for the ROM 9. In this example, the latch circuit 27 and the ROM 26 are connected in parallel.
28 and 29 are connected in series, and the output of each latch circuit is supplied to variable frequency divider circuits 1.3 and 14.15, respectively, in the same manner as above.

Signals from the control circuit 7 at each timing are simultaneously supplied to all latch circuits 27, 28, and 29. The output of the variable frequency divider circuit 13 is directly supplied to a counter 17, the output of the frequency divider circuit 14 is directly supplied to a counter 18, and the output of the frequency divider circuit 15 is directly supplied to a counter 19. Other configurations are the first
This is similar to the example.

Therefore, when the frequency division ratio data of the first sound C is latched in the latch circuit 27, the frequency division ratio data of the second sound B is R.
When output from OM26, this data is sent to latch circuit 2.
7, and the frequency division ratio data of the first tone C is latched in the latch circuit 28. Similarly, when the frequency division ratio data of the third sound A is output from the ROM 26, this data is sent to the latch circuit 27, and the frequency division ratio data of the second sound B is sent to the latch circuit 28.
Then, the frequency division ratio data of the first tone C is latched in the latch circuit 29. In this way, without the need for a selector, D/
The A converter 20 always stores the acoustic data at the beginning of each sound, the D/A converter 21 always stores the audio data a little after the beginning of each sound, and the D/A converter 22 always stores the audio data a long time after the beginning of the sound. is supplied. Other operations are the same as those described for the first embodiment.

In the third embodiment shown in FIG. 4, the selector 8 in the second embodiment is changed to a selector 30.
．． 2.3 and the D/A converter 20°21.22. Therefore, the output of CR4゜5.6 is D/A
converter 20,2].

22, respectively. Input terminals Di”, D2”, D3°° and output terminal Q1” of selector 30
, Q2'', and Q3°° are the same as in the first embodiment, and the timing signal TI allows Ql'' = DL
", Q2" = D3", Q3°"-D2°', Ql"'-D2" by the timing signal T2.

Q2" = DI", Q3" = D3°, Qbi°-D3°°, Q2°°-D2° by timing signal T3
゛.

Q3" = DI" respectively.

Therefore, when the timing signal T3 is output, the first g
The acoustic data of C is supplied from the ROM3 to the input terminal D3- of the selector 30, and then from the output terminal Q1° to the D/A
Converter, supplied to 20, D/A converted, CR4
Envelope processing is performed using the envelope signal q1 of the timing signal T3 from . The second sound B is supplied from the ROM2 to the input terminal D2- of the selector 30, and then to the output terminal Q2°.

is supplied to the D/A converter 2], subjected to D/A conversion, and subjected to envelope processing using the envelope signal q2 of the timing signal T3 from CR5. Also, the third sound A is RO
It is supplied from MI to the input terminal D1' of the selector 30, and then from the output terminal Q3- to the D/A converter 22, and the D/A
A conversion is performed, and envelope processing is performed using the envelope signal q3 of the timing signal T3 from CR6.

When a selector is disposed between the ROM and the D/A converter in this manner, it is sufficient that the selector transmits only digital signals, which is simpler than one that transmits analog signals. Other operations are the same as in the first embodiment.

Note that the envelope processing is performed using D/D as in each of the above embodiments.
This is not limited to being performed simultaneously with A conversion, but may be performed after D/A conversion.

[Effects] The audio signal synthesis circuit of the present invention having the above configuration stores a plurality of different audio data such as the first half, middle, and second half of the generated sound, and synthesizes the sound data according to the generation timing of each sound. Since an envelope signal of the platform is generated and the sound data is envelope-processed using the envelope signal, the sound continues smoothly before and after the sound synthesis, making it possible to synthesize sounds close to the natural state.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing the first embodiment of the present invention, and Fig. 2 corresponds to the generation timing of each sound, and the envelope circuit (
CR) 4, 5.6 Stars +1. lj, C
Envelope signals generated from R4, 5, and 6, connection status between the output terminal and input terminal of selector 16, and selector 8
Fig. '3 is a block diagram showing the second embodiment of the present invention,
FIG. 4 is a block diagram showing a third embodiment of the present invention, and FIG. 5 is a waveform diagram showing the state of sound synthesized by a conventional aromatic synthesis circuit. 1.2.3...Memory circuit 4.5.6...Envelope circuit (CR) 20.21
．． 22...D/A converter 8.30...Selector 23...Mixing circuit ql, q2. q3...Envelope signal or more

Claims (2)

[Claims] (1) A plurality of storage circuits each storing different acoustic data; A plurality of envelope circuits that generate envelope signals corresponding to each of the above-mentioned acoustic data in accordance with the generation timing of each of the above-mentioned acoustic data; a plurality of D/A converters that D/A convert the acoustic data from the storage circuit and perform envelope processing using envelope signals corresponding to the acoustic data from each of the envelope circuits; An acoustic signal synthesis circuit comprising: a selector for outputting to the D/A converter to which acoustic data is supplied; and a mixing circuit for synthesizing respective output signals from the respective D/A converters. (2) The acoustic signal according to claim 1, wherein the selector outputs the acoustic data from each of the storage circuits to a D/A converter to which an envelope signal corresponding to the acoustic data is supplied. Synthetic circuit.