JPH02176795A - Musical sound generating device for electronic musical instrument - Google Patents

Abstract

PURPOSE:To enable smooth waveform modification matching variation in pitch at the time of the pitch variation by reproducing a musical sound while performing cross-fading between a waveform in a range being sounded currently and a waveform in an adjacent range to be switched. CONSTITUTION:When pitch which is specified and inputted by an input means 1 is in a 1st specific range, 1st waveform data are read out of storage means ROMs 10 and 11 and when the input pitch varies to a high or low frequency range beyond the 1st sound range, the 2nd waveform data in a 2nd high or low range adjacent to the 1st range is read out. Then the output of a CPU 4 is supplied to a level controller 5 and envelope generators 8 and 9 to decrease the amplitude of the 1st waveform data gradually and also increase the amplitude of the 2nd waveform gradually, thereby switching waveform data in different adjacent ranges by cross-fading. Consequently, even if the pitch varies, the waveform modification is performed smoothly according to the variation in the pitch.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a musical tone generation device for an electronic musical instrument, and particularly to a musical tone generation device for an electronic musical instrument that can smoothly switch output musical sound waveforms in accordance with changes in pitch. Regarding equipment.

[Prior art 1] Electronic musical instruments using pulse code modulation (PGM) have been developed for some time, and in this type of electronic musical instruments, for example, a keyboard for an electronic piano, a key switch for a wind instrument, etc. , a means for inputting a designated pitch, and an input means such as a sensor such as a switch or a bender that can change the pitch of the musical tone being generated. Then, waveform information previously stored in the storage circuit is read out corresponding to the pitch specified by these input means. As a result, a musical tone having a predetermined timbre is generated.

[Problems to be Solved by the Invention] By the way, in an electronic musical instrument that records waveforms using PCM, if a musical tone is generated by associating one waveform with a wide range of pitches, The problem is that the timbre changes considerably as the pitch changes. Therefore, a method called multisampling has been used to prevent changes in tone color when pitch changes. This is an attempt to pre-store waveforms for each predetermined range (for example, one octave) in a storage circuit, and to selectively read and generate the stored waveforms according to a specified pitch during playback.

However, even with such a musical tone generation device, if the pitch is changed by an operation such as a bender during musical tone generation, the tone color will change as described above. Therefore, in order to correspond to the specified pitch, a waveform in a higher or lower range adjacent to the range of the waveform currently being generated is read out, and the waveform before the pitch changes is switched to the waveform in this adjacent range. There are possible ways.

However, the memory space that stores the waveforms for each range is independent, and if the waveform is switched by changing addresses between unrelated memories, it will become discontinuous. In reality, it was not possible to generate musical tones by connecting waveforms in adjacent ranges.

The present invention has been made in view of the above-mentioned circumstances, and an object of the present invention is to provide a musical tone generation device for an electronic musical instrument that can smoothly change the waveform in accordance with the change in pitch even when the pitch changes. .

[Means and effects for solving the problem] That is, the present invention has a plurality of waveform data stored in advance in the storage means,
The waveform data stored in the storage means is read out by address signals sequentially generated by the address signal generation means in accordance with pitches specified by input means for specifying and inputting pitches for performance. After the amplitude of the waveform data is controlled by the amplitude control means, it is synthesized by the synthesis means, and a musical tone based on the waveform data is output from the output means. When the pitch specified by the input means is within a predetermined first range, the first waveform data stored in the storage means is read, and the pitch specified by the input means is read out. When the sound changes beyond the first range to the treble side or the bass side, unlike the first waveform data, the second range on the treble side or the bass side adjacent to the first range is changed. Read out the waveform data of 2. The amplitude control means gradually decreases the amplitude of the first waveform data when the pitch changes, and adjusts the second waveform data in response to the decrease in the amplitude of the first waveform data. The amplitude is controlled to gradually increase, and waveform data of adjacent different sound ranges are cross-faded and switched.

[Example code] The present invention will be described in detail below with reference to the drawings.

FIG. 1 shows a block configuration diagram of a musical tone generation device for an electronic musical instrument according to an embodiment of the present invention, in which an input circuit 1 includes a pitch designation input section 2 and a variable pitch input section 3. There is.

The pitch specification input section 2 is, for example, a keyboard for a keyboard instrument, and an input section for inputting pitches such as corresponding switches and frets for a wind instrument, a guitar, etc., and is a variable pitch input section. Section 3 is an input section such as a switch, bender, or various sensors that can change the pitch of the musical tone being generated and input it. The pitch specification input section 2
The outputs from the 6J variable pitch input section 3 are both input to the CPU 4 as a control circuit.

The output of the CPU 4 is supplied to a level controller 5 for cross-fade controlling the amplitude level of an output waveform, which will be described later, according to a specified pitch, and is also supplied to address controllers 6 and 7, and further to envelope generators 8 and 9. . The outputs of the address controllers 6 and 7 are supplied to waveform storage circuits (waveform ROM) 10 and 11, respectively.

The envelope generators 8 and 9 generate envelope data that changes the amplitude of the musical tone to be output over time. Further, the waveform ROMs 10 and 11 store waveforms for each predetermined range (for example, for each octave). The output of the waveform ROM10 is input to the multiplier I2 together with the output of the envelope generator 8, and the multiplication result is input to the multiplier 13. This multiplier 3 is supplied with the output of the level controller 5 as well as the multiplication result of the multiplier 12. On the other hand, the output of the waveform ROMII is sent to the multiplier 14 together with the output of the envelope generator 9.
and the multiplication result is input to the multiplier 15.

The multiplier 15 is supplied with the output of the level controller 5 together with the multiplication result of the multiplier 14 via an inverter 16 with the level inverted.

The multiplication results of the multipliers 13 and 15 are supplied to latches 17 and 18, respectively, and then output to an adder 19. The result added by this adder 19 is then added to the latch 2.
0 and further supplied to the digital-to-analog converter 21. Next, this digital-to-analog converter 21
The signal converted into an analog signal is amplified by an amplifier 22 and outputted from a speaker or the like (not shown).

FIG. 2 shows, for example, an address controller 6, in which the output from the CPU 4 is coupled to a loop start address register 23, a start address register 24, a pitch register 25, a loop end address register 2G, and an end address register 27. . The output of the start address register 24 is input to the current address register 29 via a gate 28 controlled by the output from the CPU 4. The output of the current address register 29 and the output of the pitch register 25 are input to an adder 30, and the result of the addition is latched into a latch 31.
Output to. The output of this latch 31 is input to a comparator 32 along with the output of the loop end address register 26. This comparator 32 outputs "1" when the current address is smaller than the loop end address, and "1" when the current address is smaller than the loop end address.
0" output. Also, the output of latch 3■ is the comparator 3
The signal is input to the current address register 29 through a gate 33 controlled by the output of 2, and through a gate 35 which is controlled to open via an inverter 34 when the note-on signal (- generation signal) from the CPU 4 is off (low level). It is now possible to do so.

Further, the output of the comparator 32 is passed through the inverter 37 to
It is input to an AND circuit 88 together with the output of the comparator 36.

The comparator 36 compares the output of the latch 31 with the output of the end address register 27, and outputs "1" when the current address is smaller than the end address, and outputs "0" when the opposite is true.

The output from the comparator 3G via the inverter 39 is input to the AND circuit 40 together with the output from the comparator 32 via the inverter 37. And circuit 3
8 and 40 are coupled to gates 41 and 42 for controlling the input supply of the output of the loop start address register 23 and the output of the current address register 29 to the current address register 29, respectively.

Therefore, the output of the loop start address register 23 is input to the current address register 29 via the gates 41 and 35.

The details of the address controller 6 have been described above, but since the address controller 6 has the same configuration as the address controller 6, the explanation thereof will be omitted here.

FIGS. 3(a) and 3(b) show examples of waveforms stored in the waveform ROMs 10 and 11. In this case, the waveform ROMs 10 and 11 store waveform data of adjacent ranges alternately for each octave. As will be described later, waveforms 2 and 4 are stored in the waveform ROM 10, and waveforms 1 and 3 are stored in the waveform ROM 11 (see FIG. 4). Also, Fig. 3(a) and Fig. 3(b)
represents waveforms of adjacent sound ranges. In the figure, the horizontal axis represents time, and the vertical axis represents amplitude. Also, ST...
ST2, LSl LS2, LEl LK2 and ED,
ED2 has respective registers 24.23.20.
27 represents the start address, loop start address, loop end address, and end address stored in 27.

Figure 4 shows the relationship between adjacent ranges; for example, waveform 1 is the range with pitches 03 to B3, and waveform 1 is the range with pitches 03 to B3;
4 as waveform 2, pitches C5 to B5 as waveform 3, pitches C6 to B6 as waveform 4, and so on. These waveforms are used within the range specified by the low key position (LK) and the high key position (HK). Specifically, for example, as LK of waveform 1, Pitch 03 is set as HKl, and B3 is set as LK of waveform 2.
2 corresponds to pitch 04, and HK 2 corresponds to pitch B4. The same applies below. Each waveform has a start address (S'r), an end address (ED), a loop start address (LS), and a loop end address (
LE). Further, for example, between pitches B3 and 04, the output is performed by crossfading so that the levels of waveform 1 and waveform 2 intersect. That is, as the level of waveform 1 gradually increases (or decreases) by 1, the level of waveform 2 correspondingly gradually decreases (or increases by 1). In this way, a section in which the levels of adjacent waveforms change each other is defined as a cross-fade section. In this case, the crossfade section is
From the point where the level (mixing ratio) of adjacent waveforms is the same, each has a width of 50φ (cents) and is 100φ.
It is said that In the figure, for example, 100φ between pitches B3 and C4 is the crossfade section, but this is
It is not limited to 00ψ. The slope of the mixing level representing cross-fading within this 100φ is determined by the output of the level controller 5.

Next, with reference to flowchart 1 in FIG. 5, the operation of the musical tone generating apparatus for an electronic musical instrument configured as described above will be described.

First, a desired pitch is specified and input using the pitch specification input section 2. As a result, a parameter (parameter) for reading waveform data corresponding to the specified input pitch is set in the address controller 6 or 7 (step 5TI). That is, the CPU 4 determines to which waveform data among waveform data 1 to 4 the pitch specified and inputted from the pitch specification input section 2 belongs. In this case, since the waveform data of adjacent ranges are alternately stored in each octave in the waveform ROMs 10 and 11, the CPU 4 reads the waveform from which waveform ROM by inputting the specified pitch. I'll decide whether. Then,
In order to read out the corresponding waveform data, parameters corresponding to the specified input pitch are set in the address controller 6 or 7.

For example, if the specified input pitch is within the range of waveform 2, LS2 and ST2 are stored in the roux start address register 23, start address register 24, loop end address register 26, and end address register 27 of the address controller 6, respectively. , LK2 and ED
Set 2.

In addition, if the specified input pitch is within the range of waveform 1, the address controller 7 registers the loop start address register, start address register, and loop end address register in the same way as the address controller 6. and end address register,
It is assumed that LS, ST1, LEl, and EDl are set respectively. In addition, when a different pitch is specified, the respective address controllers 6 or 7 are read out in the same manner so that the address controller 6 reads the data of waveform 4 or the address controller 7 reads the data of waveform 3. , corresponding parameters are set so that waveform data of non-adjacent ranges are read out alternately every octave.

In this case, the pitch register 25 is set with the specified input pitch pitch of waveform 2 within pitches 04 to B4. Then, so that the amplitude of the waveform is maximized,
The maximum value of the level is set in the level controller 5 (step 5T2).

That is, the multiplier 13 is given the maximum value from the level controller 5, and the multiplier 15 is given the maximum value from the level controller 5.
The output of is inverted by the inverter 16 so that the minimum value (0) is given. Next, the CPU 4 outputs a musical tone generation instruction signal (note-on signal) to the address controller 6 and envelope generator 8 (step 5T3).
). Note that when the range corresponding to the designated pitch is waveform 1.3, the minimum value is set in the level controller 5 in order to output the waveform output of the waveform ROM 11 at the maximum level.

When the note-on signal is output from the CPU 4, the address con! - Inside the roller 6, the gate 28 is opened and the output ST2 from the start address register 24 is input to the current address register 29. The output of the current address register 29 is input to the adder 30 together with the output from the pitch register 25, and the addition result (CA) is output to the latch 31.

By the way, the value stored in the latch 31 while being incremented is added to the output of the current address register 29 by the adder 3.
0 and the pitch value output from the pitch register 25 is added at every predetermined timing (sampling period). The output of the latch 31, which is added by the pitch value, is supplied to the comparator 32, and returns to the current address register 29 via the gate 33 controlled by the output of the comparator 32, and further via the gate 35. . As a result of comparing the output CA of the latch 31 and the output LE2 of the loop end address register 26, if the current address CA has not reached the loop end address LE2, the gate 33 is opened by the output of the comparator 32. Further, the gate 35 is turned on immediately after the note-on signal is input, the gate 28 is opened, and the output of the start address register 24 is input to the current address register 29. As a result, the current address CA stored in the latch 31 is changed to the loop end address LE2.
latch 31, gate 33, gate 35 until reaching
, the current address register 29, and the adder 30 to form a loop and continue to advance. This current address CA is then given to the waveform ROM10 as a read address.

Since the step rate of this current address CA is determined by the output of the pitch register 25, the pitch can be controlled.

Then, when the output from the latch 31 reaches the value LE2 of the loop end address register 26, the comparator 32
The output of turns off the gate 33 and also turns off the inverter 37.
are input to one input terminal of AND circuits 38 and 40, respectively. The outputs of these AND circuits 38 and 40 are determined by the output of comparator 30. That is, the comparator 3G compares whether the current address OA held in the latch 31 has reached the output value ED2 from the end address register 27. The output is then input to the other input terminal of the AND circuit 38 and to the other input terminal of the AND circuit 40 via the inverter 39. Therefore, when the current address CA is between the loop end address LE2 and the end address ED2, the AND circuit 38 outputs a loop end signal to the gate 4.
is opened to pass the output of the loop start address register 23. At this time, since the end signal is not output from the AND circuit 40, the gate 42 is not opened. In other words, the output of the loop start address 23 is
2 is output to the current address register 29 via the gate 4 and the gate 35. Therefore, when the current address CA reaches the loop end address LE2, it returns to the loop start address LS2 and returns to the loop section (L
S2 to LE2) are formed. In this way, by repeating this loop section, the musical tone to be generated continues to be output. Therefore, for example, among the waveforms from the start address ST2 to the end address E D 2 of waveform 2 as shown in FIG. 3(b), the start address S
Data from T2 to loop end address LE2 is read,
While continuing, the loop section from the loop start address LS2 to the loop end address LE2 is repeatedly read out according to the duration time. If the tone is to be muted, data up to the end address E D 2 are read out and musical tone generation is ended. For example, in this case, the loop end address register 26 may be rewritten to the end address ED2. still,
When the current address reaches the end address ED2, a "0" output is given from the comparator 36, and the gate 4
2 is opened, and the output of the current address register 29 is
No address increments are done, only looping through gates 42.35. At this time, the gate 33 is connected to the comparator 3.
Since the output of 2 is "0", it is closed.

By the way, during musical tone generation, the waveform read from the waveform ROM10 is supplied to the multiplier 12. And CPU4
A predetermined envelope generated by the envelope generator 8 receiving the note-on signal from the multiplier 12
is supplied to As a result, the predetermined envelope generated from the envelope generator 8 is converted into the waveform RO
The waveform read from Ml0 is multiplied by multiplier 12 and output to multiplier 13. As described above, in the multiplier 13, in this case, the maximum value is multiplied by the output of the multiplier 12, and then outputted via the adder 19, latch 20, etc.

By the way, after the note-on signal is output in step ST3 mentioned above, whether or not a control signal is input by the variable pitch input section 3 to cause a change in the pitch, that is, the frequency, of the pitch currently being produced. (Step 5T4). This change in pitch is detected by the CPU 4 by monitoring the variable pitch input section 3, and in step ST4, the same determination is repeated until the pitch changes. still,
Although not shown, other processing is performed between step ST3 and step ST4 as necessary. If the pitch changes, it is determined whether the change in pitch is above or below the current pitch (step 5T5).

In this step ST5, if there is a downward pitch change, the process proceeds to step ST6, where it is determined for the first time whether or not there has been a downward pitch change. And if the change is new to you,
Proceeding to step ST7, in this case, the waveform address of the waveform data adjacent to the range of waveform 2 is set in the address controller 7.

In this case, this address controller 7 operates in the same manner as the address controller 6, and reads out the waveform data adjacent to the sound range of waveform 2. 1) Read the waveform data. And address controller 7
The waveform read from waveform ROMII by
4. Further, a predetermined envelope generated by an envelope generator 9 that receives a note-on signal from the CPU 4 is supplied to a multiplier 14 . As a result, the waveform read from the waveform ROMII is multiplied by the predetermined envelope generated from the envelope generator 9 by the multiplier 14, and then outputted to the multiplier 15. Further, in step ST6,
If the change is not the first time, the process proceeds to step ST8 without proceeding to step ST7.

In this step ST8, a change in pitch is determined. In this case, since the pitch of low key LK2 is C4, it is difficult to determine whether the pitch is between pitches B3 and 071, that is, L
K −100φ≦Pitch≦L K .

Determine. Here, if the pitch is not between pitches 83 and 04, the process proceeds to step 5TIO, which will be described later.
In the case of K2-100φ≦pitch≦LK2, that is, in the case of continuous pitch change due to portamento, pitch bend, etc., the process proceeds to step ST9, where waveform mixing is performed by cross-fade. That is, the value of the level controller 5 is updated.

That is, the value of the level controller 5 changes depending on the pitch, and in this case, the mixing ratio of waveform 1 and waveform 2 is determined depending on the pitch. Therefore, if the pitch is 0.1 or more, musical tones are generated using only waveform 2 as described above, but as the pitch changes from C61 to B3, as the pitch changes,
The level of waveform 2 changes from the maximum value to the minimum value, while the level of waveform 1 changes from the minimum value to the maximum value, thereby being cross-fade controlled. This is because the output from the level controller 5 is supplied to the multiplier 13, and
This is because the output inverted by the inverter 16 is supplied to the multiplier 15, so that when one level increases, the other level decreases in a complementary manner. As a result, the level of the musical tone of waveform 2 decreases, and the level of the musical tone of waveform 1 increases by 1, so that the waveform is switched and a natural musical tone change is performed. In this way, each time a change in pitch is detected in step ST4, the following steps Sr1,
Sr6, Sr8, and Sr9 are executed, and the series of processes is repeated until step 5TIO is satisfied.

Then, in step 5TIO, a change in pitch is determined again, and it is determined whether the pitch is equal to or less than LK2-100ψ. In this case, when the cross-fade control is performed in step ST9 and the waveform shifts to another waveform, for example, after shifting from the range of waveform 2 to the range of waveform 1,
The determination result in step 5TIO is rYEsJ, and the process proceeds to step 5TII. Of course, in this case, the determination in step ST8 is rNOJ. And this step 5TII
Then, the original address controller, in this case address controller 6, is turned off, that is, initialized.

Therefore, from now on, only the waveform] will be output with the specified pitch.

On the other hand, if it is detected in step ST5 that there is an upward pitch change, step ST5T
It is not until step 12 that it is determined whether or not there has been an upward change. If the change is for the first time, proceed to step 5T13 and set parameters in the address control rough to read out waveform data adjacent to the range of waveform 2, in this case waveform 3, and Output the node-on signal in the same way as in the case. Further, in step 5TI2, if the change is not the first time, the process does not proceed to step 5TI3 but proceeds to step 5TI4.

In this step 5T14, a change in pitch is determined.

In this case, since the high key HK2 is B4, it is determined whether the pitch is between 84 and 05, that is, HK≦pitch≦HK+100φ. Here, the pitch is B
, and C5, step 5T1B described later.
However, if HK2≦pitch≦HK2+100φ,
Proceeding to step 5T15, waveform mixing control by cross-fade is performed. That is, the level controller 5
The value of changes depending on the pitch. In this case, the mixing ratio between waveform 3 and waveform 2 is determined depending on the pitch. therefore,
As the pitch changes from B4 to 05, the level of waveform 2 changes from the maximum value to the minimum value, while the level of waveform 3 mixed by cross-fade control changes from the minimum value to the maximum value. Then, the level of the musical tone of waveform 2 decreases and the level of the musical tone of waveform 3 increases by 1, and the waveform is switched.

Also, in step 5TiB, the above-mentioned step 5TIO
The change in pitch is determined in the same way as in . In this case, it is determined whether the pitch is HK2+100φ or more. After the cross-fade process has been performed in step 5T15 and the waveform has shifted to another waveform, for example, after the waveform has shifted from the range of waveform 2 to the range of waveform 3, the process proceeds to step 5TL7.

Then, in step 5T17, the original address controller, in this case address controller 6, is turned off, that is, initialized. Therefore, from now on, only waveform 3 will be output with the specified pitch.

In this way, when the pitch of a musical note that is currently being sounded changes outside the range, the waveform of the currently sounding range and the waveform of the adjacent range that is to be switched are played back while loss fading is performed. Ru. For this reason, when the pitch changes,
Waveforms can be changed smoothly according to changes in pitch.

In the above-mentioned embodiment, it is assumed that the two waveform memory circuits store waveform data of adjacent ranges alternately for each octave, but both waveform memory circuits each store all the waveform data of each octave. It may also be one that stores waveform data of a range. In that case, for example, the address controller 6 can be operated first for any sound range.

Furthermore, the two-system musical tone generation circuit is not limited to the above-described embodiment, but can be realized by using two channels of a plurality of channels of sound source device through time-sharing processing. In short, to create one sound, it is sufficient to provide at least two musical tone generating channels, and each musical tone generating channel may have any structure.

Further, the envelope control by the envelope generator may be performed after the adder 19. In that case, it is sufficient to use one system of envelope generators, which further simplifies the configuration.

Furthermore, the present invention can be applied to things other than expressing waveforms using PCM technology. For example, a differential PCM method, an adaptive differential PCM method, etc. can be adopted.

[Effects of the Invention] Since the present invention is configured as described below, it produces the effects described below.

When the pitch specified by the input means is within a predetermined first range, the address signal generation means reads first waveform data stored in the storage means, and reads out the first waveform data stored in the storage means, When the pitch exceeds the first range and changes to the treble side or the bass side, the first
The amplitude control means reads second waveform data of a second sound range on the high or low sound side adjacent to the first sound range, and the amplitude control means changes the pitch specified by the input means. cross-fade control is performed to gradually decrease the amplitude of the first waveform data and gradually increase the amplitude of the second waveform data in accordance with the decrease in the amplitude of the first waveform data. Therefore, even if the pitch changes, the waveforms can be naturally connected.

Then, the address signal generation means reads out the waveform data stored in the storage means from the beginning, so that
Control becomes easier.

Furthermore, the storage capacity is reduced by the storage means having first and second storage means, and waveform data of non-adjacent ranges being alternately stored in the first and second storage means, respectively. be able to.

[Brief explanation of the drawing]

FIG. 1 is a block configuration diagram of an embodiment of the present invention, FIG. 2 is a block configuration diagram of the address controller in FIG. 1,
Figures 3 (a) and (b) are waveform diagrams of the embodiment of the present invention, Figure 4 is a diagram showing the relationship between tonal ranges to explain the cross-fade in this embodiment, and Figure 5 is a diagram of the present invention. It is a flowchart showing the operation of the embodiment. 1... Input circuit, 4... CPU, 5... Level controller i-roller, 6.7... Address controller, 1
0.11... Waveform storage circuit (waveform ROM), 12.
13.14.15... Multiplier, 16... Inverter, 19... Adder. Applicant's agent Patent attorney Takehiko Suzue

Claims (3)

[Claims] (1) An input means for inputting a specified pitch for a performance, a storage means for storing a plurality of waveform data in advance, and an input means for inputting an address signal for reading out the waveform data stored in the storage means. address signal generation means that sequentially generates address signals according to pitches specified by the address signal generation means; and amplitude control means that controls the amplitude of the waveform data stored in the storage means;
comprising a synthesizing means for synthesizing waveform data read from the storage means and whose amplitude is controlled by the amplitude controlling means, and an output means for outputting a musical tone based on the waveform data synthesized by the synthesizing means, The address signal generating means reads first waveform data stored in the storage means when the pitch specified and inputted by the inputting means is within a predetermined first range; When the input pitch exceeds the first range and changes to the treble side or the bass side, unlike the first waveform data, the second waveform data adjacent to the first range is on the treble side or the bass side. reads out second waveform data in a musical range, and the amplitude control means gradually decreases the amplitude of the first waveform data when the pitch specified by the input means changes, and the amplitude control means gradually decreases the amplitude of the first waveform data, and A musical tone generation device for an electronic musical instrument, characterized in that cross-fade control is performed so that the amplitude of the second waveform data is gradually increased in accordance with a decrease in the amplitude of the data. (2) The musical tone generation device for an electronic musical instrument according to claim 1, wherein the address signal generation means reads out the waveform data stored in the storage means from the beginning. (3) The storage means has first and second storage means, and waveform data of non-adjacent ranges are stored in the first and second storage means, respectively. A musical tone generation device for an electronic musical instrument according to item 1.