JPH0429080B2 - - Google Patents

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a musical tone signal generating device, and more particularly to a waveform memory readout device that generates a sound signal by reading a waveform memory that stores musical waveform information. This invention relates to a method for improving a musical tone signal generating device.

[Prior art]

Conventionally, as a musical tone signal generating apparatus for a waveform memory reading method, there is one described in, for example, US Pat. No. 3,515,792. In this device,
Since musical tone signals are generated by repeatedly reading out a waveform memory that stores musical waveforms for one period, the waveform of the generated musical tone signals is always the same (the same musical waveform is repeated). For this reason, the musical tone signal becomes extremely monotonous, making it impossible to obtain a complex musical tone signal whose timbre changes over time, such as the sound of a natural musical instrument.

Therefore, in order to improve this point, the acoustic waveform for multiple cycles in the attack part of a musical tone and the acoustic waveform in the sustain part (sustain part) are stored in a waveform memory, and a musical tone generation start command is generated (pressed). key) to sequentially read out the musical waveforms of the attack section,
Subsequently, a musical tone signal generating device has been proposed in which a musical tone signal is generated by repeatedly reading out the musical sound waveform of the sustain section until a musical tone generation end command is issued (key release) (US Pat. No. 3,539,701). ). According to this improved musical tone signal generator,
It is possible to generate a musical tone signal whose timbre etc. change over time.

[Problems that the invention attempts to solve]

However, in the case of this improved device, reading of the waveform memory completely stops when a musical tone generation end command is generated (key released), so no musical tone signal is generated after the key is released. There is a problem in that it is not possible to express the lingering attenuated pronunciation that is important in natural instruments such as.

This invention has been made in view of the above points, and is a high-quality product whose timbre changes over time from the start to the end of musical sound generation, and which can effectively express the lingering decay of natural musical instruments. It is an object of the present invention to provide a musical tone signal generating device that can easily generate musical tone signals with a simple configuration.

[Means for solving problems]

According to the musical tone signal generating device according to the present invention, the first waveform memory portion sequentially stores information about a plurality of cycles of the waveform in the attack portion of the musical tone in each address, and the information about the waveform for the plurality of cycles of the musical tone. A waveform memory consisting of a second waveform memory part stored in each address sequentially, and each address of the first waveform memory part sequentially stored at a predetermined speed from the initial address upon generation of a musical tone generation start command. The stored waveform information is sequentially read out, and after reading out the first waveform memory section, each address of the second waveform memory section is sequentially and repeatedly designated at a predetermined speed and stored in each address. waveform memory reading means for sequentially and repeatedly reading out the waveform information read out; a decay envelope adding means for adding the first decay envelope or the second decay envelope, the waveform information read from the waveform memory and the waveform to which the decay envelope is added by the decay envelope adding means. A musical tone signal is generated based on the information.

Further, according to the musical tone generating device according to the second aspect of the invention, information about the waveform for a plurality of cycles in the musical tone attack portion is sequentially stored in each address.
The second waveform memory part and information about the waveform for multiple periods of musical tones are sequentially stored in each address.
a waveform memory consisting of a waveform memory section, and each address of the first waveform memory section is sequentially designated from the initial address at a predetermined speed upon generation of a musical tone generation start command, and the waveform information stored in each address is sequentially transferred. After reading out the first waveform memory portion, the first waveform memory portion is attenuated relatively slowly before each tone generation end command is generated in the second waveform memory portion.
, and generates a second decay envelope that attenuates more rapidly than the first decay envelope in response to generation of a musical tone generation end command, sequentially at a predetermined speed. and a waveform memory reading means for sequentially and repeatedly reading out the waveform information stored in each address with repeated designation;
a decay envelope adding means for adding a decay envelope to the waveform information read by the waveform memory reading means; a dump command generating means for selectively generating a dump command; when the dump command is being generated; The decay envelope adding means is controlled so as to rapidly attenuate the decay envelope in response to generation of a musical sound generation end command, and when the dump command is not generated, the decay envelope is control means for controlling the decay envelope addition means to add a decay envelope, the waveform information read from the waveform memory and the information of the waveform to which the decay envelope has been added by the decay envelope addition means; The musical tones are pronounced based on the following.

Furthermore, according to the musical tone generating device according to the third aspect of the invention, the first waveform memory portion sequentially stores information about a plurality of cycles of waveforms in the attack portion of the musical tone in each address, and the plurality of waveforms in the sustain portion of the musical tone. A waveform memory consists of a second waveform memory section that sequentially stores information about the waveform in each address, and upon generation of a musical sound generation start command, each address of the first waveform memory section is sequentially stored at a predetermined speed from the initial address. Sequentially reading out the information of the waveform stored in each specified address,
After reading the first waveform memory portion, the second waveform memory portion is read out.
waveform memory reading means for sequentially and repeatedly specifying each address of the waveform memory portion of the waveform memory portion at a predetermined speed and sequentially and repeatedly reading out the waveform information stored in each address; and generation of a tone generation termination command. Decay envelope adding means for adding a decay envelope to the waveform information read from the second waveform memory portion in response to
From generation of the musical sound generation start command to generation of the musical sound generation end command, the waveform information of the sustain part is repeatedly output following the waveform information of the attack part, and after the generation of the musical sound generation end command is generated, the waveform information of the sustain part is repeatedly output. The waveform information of the sustain portion to which the K-envelope is added is output, and a sustained tone musical tone signal can be generated based on the output waveform information.

[Effect]

According to the first invention, the attack part and the musical tone part are taken out from separate memory parts, a decay envelope is added by the decay envelope adding means, and the generation of the decay envelope is controlled by the occurrence of the end of musical tone generation. Because it is
It is possible to generate a complex musical tone signal to which envelopes with different attenuation characteristics are added before and after the generation of a musical tone generation end command.

Further, in the second aspect of the invention, the damping characteristic can be selected by providing a dump command generating means.

Furthermore, in the third invention, the stored waveforms are shared for the sustain section and the decay section, which have relatively few timbres, making it possible to produce attenuated sound with a small amount of waveform memory.

〔Example〕

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

FIG. 1 shows a first embodiment of a musical tone signal generating device to which the present invention is applied, and is suitable for generating percussive musical tone signals, but is of course not limited to this. .

According to this embodiment, each address
It has three types of waveform memories WM ₆₁ , WM ₆₂ , and WM _{63 addressed by AD 61 , AD 62 , and} AD ₆₃ .
The first waveform memory WM ₆₁ stores information on multiple cycles of the musical sound waveform of the attack portion, and the second waveform memory WM ₆₂ stores information on multiple cycles of the necessary musical sound waveform. Third waveform memory
WM ₆₃ stores the envelope waveform of Decay. Therefore, after reading the first waveform memory WM ₆₁ , if envelope shaping is performed by the third waveform memory WM ₆₃ while repeatedly reading the second waveform memory WM ₆₂ , it is possible to perform envelope shaping using a simple memory. A musical tone signal having an attack part and a decay part can be generated. Here, it is optional whether or not the third waveform memory WM ₆₃ has a sustain envelope. In this embodiment, it is assumed that an attack envelope is added to the multi-period waveform of the attack portion stored in the waveform memory WM ₆₁ .

Next, the structure and operation of this example will be clarified while explaining the process by which musical tone signals are generated by this example.

FIG. 2 shows a keyboard device that generates various control signals (musical tone generation start command signal, musical tone generation end command signal) for controlling the generation of musical tone signals when the above musical tone signal generation device is used in a keyboard electronic musical instrument. Since this keyboard device is used in the musical tone signal generating device shown in FIG. 1, this keyboard device will first be explained. In the figure, the circuit for one key switch KSW is shown for convenience of explanation, but it goes without saying that in reality, a similar circuit may be adopted for all or some of the keys on the keyboard. be.

When the key switch KSW is closed or opened by key operation, the key switch is closed or opened by the action of power supply E.
A key press signal A as shown in FIG. 3A appears at one end of the KSW in response to opening and closing of the key switch KSW. In this device, in addition to using this signal A as it is, an inverter INV4 inverts the key press signal A to form an inverted key press signal (FIG. 3B).

The key depression signal A thus generated is converted by a differentiating circuit comprising R ₁ and a capacitor C ₁ into pulse signals having opposite polarities when the key is depressed and when the key is released. Among these pulse signals, the pulse signal corresponding to the key release is cut by the diode _D1 to obtain a key press pulse KD (musical sound generation start command signal) as shown in FIG. 3C. Also, this pulse
KD is inverted by the inverter INV ₁ to form an inverted key press pulse (FIG. 3 D).

In addition, the inverted key press signal obtained by inverting the key press signal A by the inverter INV ₂ is sent to the capacitor C ₂ and the resistor.
A differentiating circuit consisting of R ₂ obtains pulse signals corresponding to the key press and key release in the same way as obtained by the above-mentioned differentiator circuit (C ₁ , R ₁ ), and a pulse signal corresponding to the key press. is cut by diode _D2 to obtain a key release pulse KR (musical tone generation end command signal) as shown in FIG.

When the key press pulse KD explained in FIG. 2 is generated by a key press operation, the flip-flop FF ₆₁ is set and continues to send out a Q output of "1". Therefore, a clock pulse φ with a predetermined period is sent to the addresser AD ₆₁ via the AND circuit AND ₆₁ .
The addresser _AD61 addresses the aforementioned waveform memory _WM61 and sequentially reads out the waveform of the attack portion stored in the memory _WM61 . The waveform read from the waveform memory _WM61 is outputted as a musical tone signal via the adder _SM60 . When addresser AD ₆₁ sends out the last bit output, flip-flop FF ₆₁
is reset, and as a result, the clock pulse φ is no longer input to the addresser _AD61 , and reading from the waveform memory _WM61 is completed.

Here, the addresser AD ₆₁ can be configured with a counter and a decoder, for example, as shown in FIG.
Cleared before counting starts. Similar configurations can be applied to other addressers described in this specification. The waveform memory WM ₆₁ can be configured using a ROM or the like, and the other waveform memories described in this specification can have a similar configuration.

Next, when the reading of the first waveform memory WM ₆₁ that stores waveforms for multiple cycles of the attack portion is completed and the final bit output "1" of the addresser AD ₆₁ is sent out, this output is sent to the flip-flop FF _61. 2nd and 3rd waveform memory
Addresser AD ₆₂ that addresses WM ₆₂ , WM ₆₃ ,
Also used as signal 1MF to drive _AD63 .

This signal 1MF sets the D-type flip-flop FF ₆₂ via the OR circuit OR ₆₁ , and the output of the flip-flop FF ₆₂ is self-held under the input conditions of the AND circuit AND ₆₂ . The AND circuit is activated by the output “1” of flip-flop FF _62.
The AND ₆₃ condition is satisfied, and this AND circuit AND ₆₃
The addresser AD ₆₂ is driven by a clock pulse φ with a predetermined period through the waveform memory WM ₆₂ .
The contents of are read out. Here, the AND circuit
The input conditions for the output of AND ₆₂ to be “1” are:
The inverted key press signal is “1” and the addresser AD ₆₃ addresses the third waveform memory WM ₆₃ .
The inverted output of the final bit output DF by the inverter INV ₆₂ is "1". Therefore, if the reading of the third waveform memory WM ₆₃ is not completed after the key is pressed, the AND condition of the AND circuit AND ₆₂ is satisfied, and the flip-flop FF ₆₂ holds itself.

Also, a D-type flip-flop provides a signal for driving the addresser AD ₆₃ under the same signal conditions as the flip-flop FF ₆₂ is self-holding.
FF ₆₃ is self-maintained by the cooperation of OR circuit OR ₆₂ and AND circuit AND ₆₄ .

However, the addresser AD ₆₃ that addresses the third waveform memory WM ₆₃ is given a drive signal when the input condition of the AND circuit AND ₆₅ is satisfied. One of the input conditions of this AND circuit AND ₆₅ is the output of the aforementioned self-holding flip-flop FF ₆₃ ,
The other is the Decay command signal DY, which is formed as described below.

There are three types of this Decay command signal DY. First, the key is in the depressed state and the key depression signal A (Figure 3 A)
is occurring, the AND circuit is activated by the clock signal _φL with a relatively long clock cycle.
AND The AND condition of ₆₆ is satisfied. Therefore, the addresser AD ₆₃ reads the data from the third waveform memory WM ₆₃ at a relatively slow speed corresponding to the clock signal _φL .
A relatively gentle decay envelope waveform is added to the waveform addressed from the second waveform memory WM ₆₂ in the multiplier MU ₆₀ and output via the adder SM ₆₀ . .

Second, if the key is not in the depressed state and the inverted key depression signal (Fig. 3B) is generated and the damper pedal DP is not depressed, the AND condition of the AND circuit _AND68 is satisfied, and the above first The waveform read out from the waveform memory WM ₆₂ by the clock signal φ _L as in the case of
An envelope is added. Thirdly, when the damper pedal DP is closed, the output of the inverter INV ₆₁ becomes "1", and if the key is not in the pressed state and an inverted key pressed signal is generated, the AND condition of the AND circuit AND ₆₇ is satisfied and relatively Short period clock signal φ _H is transferred to addresser AD ₆₃ via OR circuit OR ₆₃ . For this reason, the addresser AD ₆₃ addresses the third waveform memory WM ₆₃ at a relatively fast speed and therefore the waveform read out from the second waveform memory WM ₆₂ is addressed relatively quickly in the multiplier MU ₆₀ . Or add a sudden decay envelope waveform,
It is sent out as a musical tone signal of the decay part via an adder _SM60 . When reading out the Decay envelope waveform from the third waveform memory WM ₆₃ is completed,
Since the final bit output signal DF of addresser AD ₆₃ is “1”, this signal DF is inverted.
The inverted signal inverted by INV ₆₂ becomes “0”,
The AND condition of AND circuit AND ₆₄ no longer holds true. As a result, flip-flop FF ₆₃ is reset and its Q output becomes "0", so that
When the clock pulse φ _L or φ _H is no longer input to the addresser AD ₆₃ , the reading of the waveform memory WM ₆₃ is completed and the generation of musical tone signals is stopped.

The above inverted signal is also input to the AND circuit AND ₆₂ , and the flip-flop is activated when the reading of the waveform memory WM ₆₃ is completed (that is, the musical tone signal generation is completed).
I also try to reset FF ₆₂ . As a result, reading of the waveform memory WM ₆₂ is also stopped. The reason why the inverted key press pulse is input to the AND circuits AND ₆₂ and AND ₆₄ is to reset the flip-flops FF ₆₂ and FF ₆₃ when a key is pressed (at the start of musical tone signal generation). In this case, if the key press operation is performed again while the waveform memory WM ₆₃ is being read in the key release state (before the signal DF is generated), the inverted key press pulse will become "0". This resets flip-flops FF ₆₂ and FF ₆₃ . This ensures that there is no problem in generating musical tone signals when a key is pressed again.

Although the above embodiment shows only a circuit corresponding to one key, it goes without saying that the same type of configuration may be adopted for all keys or some keys of an electronic musical instrument. .

As can be seen from the above, according to the first embodiment, the multi-cycle waveform of the attack portion is read out by the first waveform memory WM ₆₁ immediately after the key is pressed, and subsequently, the multi-cycle waveform of the attack portion is read out by the first waveform memory WM 61. To the multi-cycle waveform repeatedly read out from WM ₆₂ , (a) if the damper pedal DP is not operated, a gentle decay envelope is added regardless of whether the key is pressed or released, and (b) the damper pedal DP is operated. A rapid decay envelope can be added immediately after the key is released.

Therefore, according to this embodiment, the waveform memory
By storing waveforms for multiple cycles in WM ₆₁ and WM ₆₂ (of course, the waveforms for each cycle can be different), musical tones in which the timbre, etc. changes over time from the start to the end of musical tone generation are created. Can generate signals. Furthermore, by appropriately operating the damper pedal DP, a musical tone signal can be obtained in which an effect similar to a piano's damper effect is imparted in the decay portion.

FIG. 5 shows a second embodiment of the musical tone signal generating device according to the present invention, in which the musical tone signal is attenuated without using a damper pedal.

According to this embodiment, each address
It is equipped with three types of waveform memories _WM71 , _WM72 , _{and WM73} that are addressed by AD71, _{AD72 ,} and _AD73 .
The first waveform memory WM ₇₁ stores musical waveform information for multiple cycles of the attack part to which the attack envelope has been added, and the second waveform memory
WM ₇₂ stores information for a plurality of cycles of the necessary musical sound waveform, and a third waveform memory WM ₇₃ stores the sustain or decay envelope waveform following attack. Therefore, after reading the first waveform memory WM ₇₁ , the second waveform memory WM ₇₂ is repeatedly read, and the envelope waveform read from the third waveform memory WM ₇₃ in response to the key release is applied to the multiplier. In MU ₇₀ , it is added to the output of the second waveform memory WM ₇₂ , and the adder
Send musical tone signals from SM ₇₀ .

Next, the structure and operation of this example will be clarified while explaining the process by which musical tone signals are generated by this example.

Flip-flop FF ₇₁ , AND circuit AND ₇₁ ,
And due to the structure of the addresser AD ₇₁ , the multi-cycle waveform of the attack part of the musical tone is read out from the first waveform memory WM ₇₁ upon the arrival of the key press pulse KD, and the final bit output signal of the addresser AD ₇₁ is sent to the flip-flop FF _71. as a reset signal, and as a second
Addresser that addresses the waveform memory WM ₇₂ of
Regarding the point that the AD ₇₂ activation signal is 1MF, see Part 1.
This is the same as the embodiment, and the explanation will be omitted.

Furthermore, when reading the second waveform memory WM ₇₂ , the OR circuit is activated by the signal 1MF.
A D-type flip-flop FF ₇₂ is set via OR ₇₁ , and the output of the flip-flop FF ₇₂ is self-held under the input conditions of the AND circuit _{AND 72 . 72} is driven, and the contents of the second waveform memory WM ₇₂ are read out. Here, the input conditions of the AND circuit AND ₇₂ are as in the AND circuit AND ₆₂ of the first embodiment, the inversion of the inverted key press pulse and the final bit output signal DF of the third addresser AD ₇₃ by the inverter INV ₇₀ . formed by the signal.

The third waveform memory WM ₇₃ is read by setting the D-type flip-flop FF 73 via the OR circuit OR ₇₂ in response to the key release pulse KR, and by setting the D-type flip-flop FF ₇₃ via the OR circuit OR 72.
Under the input condition of AND ₇₄ , the output of flip-flop FF ₇₃ is self-held, and via the AND circuit AND ₇₅ ,
This is done by driving addresser _AD73 with clock signal CK. That is, when the key is released and a key release pulse KR is generated, this pulse KR sets the flip-flop FF ₇₃ via the OR circuit OR ₇₂ , and the input condition of the AND circuit AND ₇₄ is the AND regarding the second waveform memory WM ₇₂ . Since it is similar to that in circuit AND ₇₂ , the output of flip-flop FF ₇₃ is self-holding. Thus, the AND circuit
Since there is an input on one side of AND ₇₅ , it is an AND circuit.
The input condition of AND ₇₅ is met, and the addresser AD ₇₃ executes the address in accordance with the period determined by the clock signal CK, and reads out the contents of the waveform memory WM ₇₃ .
As can be seen from the above, the clock signal CK determines the decay speed and may be set arbitrarily. Further, the decay ends when the addresser AD ₇₃ sends out the final bit output signal DF, and this final bit output signal DF is inverted by the inverter INV ₇₀ to become an inverted signal indicating the end of the decay. This signal is an AND circuit
Since one of the inputs of AND ₇₂ and _{AND 74} is set to "0", the AND condition of the AND circuits AND 72 and AND ₇₄ is broken, and the input of the second and third addressers AD ₇₂ and AD ₇₃ disappears, so the second and third waveform memory
Reading of WM ₇₂ and WM ₇₃ is completed.

In summary, according to the second embodiment, immediately after a key is pressed, the multi-cycle waveform of the attack portion is read out from the first waveform memory WM ₇₁ and output as a musical tone signal of the attack portion via the adder SM ₇₀ . , Subsequently, the contents of the second waveform memory WM ₇₂ , which stores a multi-period waveform without envelope formation, are repeatedly read out to sustain the musical tone signal.
form part. If there is no key release operation, the output of the second waveform memory WM ₇₂ is sent to the multiplier MU ₇₀ and the adder.
When a key release pulse KR is generated by a key release operation, _{the waveform read out from the second waveform memory WM 72 in the multiplier MU 70 is} added to the waveform in the third waveform memory WM _73. A decay envelope, which is the memory content, is added, causing the generation of musical tone signals to decay and disappear.

Thus, according to this second embodiment, the first
The attack part of the musical tone signal is stored in the waveform memory WM ₇₁ .
The second waveform memory WM ₇₂ stores the sustain part of the musical tone signal, and the second and third waveform memories WM ₇₂ and WM ₇₃ store the sustain part of the musical tone signal.
A combination of these forms the decay part of the musical tone signal.

Therefore, in this second embodiment as well, by storing a plurality of periodic waveforms in the waveform memories WM ₇₁ and WM ₇₂ , it is possible to generate a musical tone signal whose timbre etc. change over time as in the first embodiment. .
Moreover, according to this second embodiment, the waveforms stored in the waveform memory WM ₇₂ are shared for the sustain and decay parts of musical tones that have relatively little timbre change. It is possible to generate a sustained-tone musical tone signal having a small number of waveform memories. Furthermore, the combination of waveform memories WM ₇₂ and WM ₇₃ can generate a musical tone signal that can effectively express the attenuated sound of a natural musical instrument after a key is released.

〔Effect of the invention〕

As described above, according to the present invention, by controlling the decay envelope according to the generation of the musical sound generation end command, envelopes with different pathos characteristics are given to the decay part before and after the generation of the musical sound generation end command. It can generate complex musical sound signals, and can produce effects similar to the damper effect on a piano.

Further, according to the second invention, both the complex relief characteristic according to the second invention and the normal relief characteristic can be easily obtained by the selectively operated tamping command generating means.

Furthermore, according to the third aspect of the present invention, the sustain and decay parts of musical tones with relatively little timbre change share the stored waveforms in the waveform memory, so that the sustain part and the decay part of musical tones with relatively little timbre change are shared. system musical tone signals can be generated using a small-capacity waveform memory. Also,
By combining the two waveform memories, it is possible to generate a musical tone signal that can effectively express the melancholy pronunciation of a natural musical instrument after the key is released.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a first embodiment of a musical tone signal generating device according to the present invention, and FIG. 2 is a keyboard device for controlling the generation of musical tone signals when the musical tone signal generating device is used in an electronic musical instrument. FIG. 3 is a waveform diagram showing an output signal of the device shown in FIG. 2, FIG. 4 is a block diagram showing a specific example of the addresser in the device shown in FIG. 1, and FIG. FIG. 6 is a block diagram showing a second embodiment of the musical tone signal generating device. WM ₆₁ , WM ₆₂ , WM ₆₃ , WM ₇₁ , WM ₇₂ ,
WM ₇₃ ...Waveform memory, AD ₆₁ , AD ₆₂ , AD ₆₃ ,
AD ₇₁ , AD ₇₂ , AD ₇₃ …Addressa, MU ₆₀ ,
MU ₇₀ ...multiplier, SM ₆₀ , SM ₇₀ ...adder, DP...damper pedal.

Claims (1)

[Scope of Claims] 1. A first waveform memory portion that sequentially stores information about a plurality of cycles of waveforms in the attack portion of a musical tone at each address, and a first waveform memory section that sequentially stores information about a plurality of cycles of waveforms of a musical tone at each address. a waveform memory consisting of a second waveform memory section in which a musical tone generation start command is issued, and each address in the first waveform memory section is sequentially designated from the initial address at a predetermined speed and the waveform is stored in each address. After reading out the information in the first waveform memory section, each address in the second waveform memory section is sequentially and repeatedly specified at a predetermined speed, and the waveform information stored in each address is sequentially read out. and a waveform memory readout means for repeatedly performing reading processing; and generating a first decay envelope that decays relatively slowly before generation of a musical tone generation termination command, and generating a first decay envelope that decays relatively slowly in response to generation of a musical tone generation termination command. Decay envelope generating means for generating a second decay envelope that attenuates more rapidly than the first decay envelope; a decay envelope addition means for adding the first decay envelope or the second decay envelope; the waveform information read from the waveform memory and the waveform to which the decay envelope is added by the decay envelope addition means; A musical tone signal generating device that generates musical tone signals based on information. 2. A first waveform memory section that sequentially stores information about waveforms for multiple periods in the attack portion of a musical tone at each address, and a second waveform memory section that sequentially stores information about waveforms for multiple periods of a musical tone at each address. upon generation of a musical tone generation start command, sequentially designates each address of the first waveform memory section from the initial address at a predetermined speed to sequentially read out the waveform information stored in each address; After reading out the first waveform memory section, each address in the second waveform memory section is sequentially and repeatedly specified at a predetermined speed, and the waveform information stored in each address is sequentially and repeatedly read out. a waveform memory reading means for executing the waveform memory reading means; a decay envelope adding means for adding a decay envelope to the waveform information read by the waveform memory reading means; a dump command generating means for selectively generating a dump command; and a dump command generating means for selectively generating the dump command. is being generated, the decay envelope adding means is controlled so as to rapidly attenuate the decay envelope in response to the generation of the musical sound generation end command, and when the dump command is not being generated, the musical sound generation end command is controlled. control means for controlling the decay envelope adding means to add the decay envelope regardless of generation of a termination command, the waveform information read from the waveform memory and the decay envelope adding means A musical tone signal generator that generates musical tones based on waveform information to which an envelope is added. 3. A first waveform memory section that sequentially stores information about a plurality of cycles of waveforms in the attack portion of a musical tone in each address, and a second waveform memory section that sequentially stores information about a plurality of cycles of waveforms in the sustain section of a musical tone in each address. a waveform memory consisting of a waveform memory section; and upon generation of a musical tone generation start command, each address of the first waveform memory section is sequentially designated from the initial address at a predetermined speed, and the information of the waveform stored in each address is read. After reading the first waveform memory portion, each address of the second waveform memory portion is sequentially and repeatedly specified at a predetermined speed, and the waveform information stored in each address is sequentially and repeatedly read. A waveform memory reading means for executing a reading process, and a decay envelope adding means for adding a decay envelope to the waveform information read from the second waveform memory portion in response to generation of a musical tone generation end command. From generation of the musical sound generation start command to generation of the musical sound generation end command, the waveform information of the sustain part is repeatedly output following the waveform information of the attack part, and after the generation of the musical sound generation end command, the waveform information of the sustain part is repeatedly output. outputting waveform information of the sustain portion to which the decay envelope is added;
A musical tone signal generating device capable of generating a sustained tone musical tone signal based on the output waveform information.