JPS6095499A - Generator for non-cyclic musical sound signal - 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 (Industrial Application Field) The present invention is applicable to, for example, an aperiodic musical tone signal generator used in electronic musical instruments.8 (Prior Art and Problems) Electronic musical instruments is generally a sound source circuit (1.-generator)
Each part, such as the opening/closing circuit (keying circuit), waveform shaping circuit (tone filter) (or waveform synthesis circuit (coupler)), etc. It is well known that

By the way, the musical sounds of stringed instruments, wind instruments, etc. have waveforms that continue in a regular and constant cycle. Therefore, among the various musical sound signals generated by electronic musical instruments, musical sound signals corresponding to the musical sounds of stringed instruments, wind instruments, etc. (musical sound signals corresponding to musical sounds in the narrow sense...periodic musical sound signals) can be easily generated by electronic musical instruments configured as described above, but for example, drums, cymbals, various drums, Sounds whose height cannot be perceived, such as the sounds of percussion instruments such as
In other words, it is difficult to produce instrumental sounds that are not included in musical sounds in the narrow sense (non-periodic musical sounds) using electronic musical instruments with the above-mentioned configuration. , bass drums, snare drums, cymbals, and other percussion instruments were used for percussion effects that were very different from the actual instrument sounds.

However, as the performance of electronic musical instruments has improved, the sound of percussion instruments has become noticeably inferior to that of other musical instruments, and efforts have been made to improve this.
In recent years, musical sound signals corresponding to the sounds of actual percussion instruments have been encoded into semiconductor memory! li! An electronic musical instrument that turns on at 11 sa.ljy, reads it out, and uses it is now available.

However, sounds of percussion instruments such as drums, cymbals, various drums, etc., whose pitch cannot be perceived, are 4r instrumental sounds (non-periodic musical sounds) that are included in musical sounds in the narrow sense.
is completely different from the 11111 period musical tones, such as those from stringed instruments and wind instruments, which consist of a fundamental tone and overtones. The sound components of each part up to the end of pronunciation indicated by Lf3 in the figure are different for every part (113 points [e
Frequency spectrum at each point in time up to '=>'6
In order to reproduce aperiodic musical tones with high fidelity,

The starting point of pronunciation indicated by tl in Figure 1 is (1, first
It is necessary to memorize the entire pronunciation (Fl) at the end of the pronunciation indicated by to in the figure and reproduce it, but to realize such a thing, a large amount of memory is required. of memory, making an electronic musical instrument with satisfactory performance extremely expensive.

Therefore, attempts have been made to use small-capacity semiconductor memory to store only part of the aperiodic musical tones, and to read and use them, but the aperiodic musical tones generated in this case Since the duration is short, it is not possible to generate percussion instrument sounds for a long time, and it is clear that an electronic musical instrument with sufficient performance cannot be provided.

In order to solve the above problem, a small capacity semiconductor memory is used to store only part of the non-periodic musical tone.
It was thought to generate a non-periodic musical sound signal over a long period of time by simply reading it out repeatedly, but it was also possible to generate a non-periodic musical sound signal over a long period of time. Like, a sound whose height cannot be felt,
In other words, since instrumental sounds that are not included in musical sounds in the narrow sense are non-periodic musical sounds, even if you extract a part of them and play them repeatedly on Jll, the generated sounds will not be periodic sounds. Therefore, it cannot be reproduced as non-periodic music 1N (sounds of percussion instruments) such as actual drums, cymbals, various drums, etc.

However, small! 1 by storing part of an aperiodic musical tone in an 11v1 semiconductor memory? Since it is extremely important for electronic musical instruments to be able to obtain good percussion instrument sounds by reading out the sound repeatedly and repeatedly, the applicant's company has made various efforts to achieve this goal. I have conducted extensive experimental research.

Then, in the experimental research, if the reading order on time axis 1 is repeated and reversed every time, the result is that the musical sound is essentially non-periodic to the auditory sense. Based on this fact, the applicant's association 11 first identified the first (t' part) corresponding to the beginning of an aperiodic musical tone and the above-mentioned aperiodic sound. A period longer than the period of the lowest frequency in the audible frequency band for a signal corresponding to the beginning part of a musical tone.
11, and a storage device in which the second signal portions taken out as 7-11 are sequentially stored, and the first signal portions stored in the storage device are read out in the order in which they are stored. The second signal portion stored in the device is read out using two readout modes: a first readout mode in which the second signal portion is read out in accordance with the storage order, and a second readout mode in which the second signal portion is read out in the reverse order of the storage order. An apparatus for generating an aperiodic musical tone signal is proposed, comprising means for alternately reading out the tone signal.

In the previously proposed aperiodic musical tone signal generation device, in order to store only a part of the aperiodic musical tone signal, the first signal portion corresponding to the beginning of the aperiodic musical tone is , and a second signal portion extracted as having a time width longer than the period of the lowest frequency in the audible frequency band for the signal corresponding to the portion other than the head of the aperiodic musical tone. What makes you remember.

This is because the sound structure of the initial monotemporal part of pronunciation is completely different from the sound structure of the other parts, and the length of the second signal part mentioned above can be set within the 8-audible frequency band. The reason why the length of the second signal portion to be repeatedly reproduced is equal to the period of the lowest frequency in the 11 frequency bands is to have a longer time width than the period of the lowest frequency in the frequency band. If also has a short time span, as a result of the repetition of 11j, instead of an aperiodic musical tone, a sound with an audible frequency corresponding to the repetition period of () will be recognized. This is because it becomes like this.

Then, as the length of +111 at the time of the second signal portion mentioned above, when the second signal portion is repeatedly illuminated and generated 11#,
Needless to say, the time length should be selected so that unstable pitch sensation (so-called wah-like pitch sensation) is minimized.

According to the previously proposed aperiodic musical sound generation device, the short-term aperiodic musical tone fllF1 stored in a small-capacity semiconductor memory generates a long-term aperiodic sound signal. It also made it possible to generate musical tone signals relatively well.

Next, regarding the outline of the operation of generating an aperiodic musical tone signal by the previously proposed aperiodic musical tone signal generating device,
The explanation will be as follows with reference to the waveform diagrams shown in FIGS. 1 and 2.

First of all, Figure 1 shows the sounds of percussion instruments such as taiko drums, cymbals, various drums, etc., whose pitch cannot be perceived, that is, instrumental sounds that are not included in musical sounds in the narrow sense (non-periodic musical sounds). This is a waveform diagram schematically showing a corresponding aperiodic musical tone signal (the envelope of the waveform is shown as a straight line for simplicity of illustration). to time te.

In the aperiodic musical tone signal shown in the first diagram, the signal portion from time t1 to time t2 is a first signal portion corresponding to the head of the aperiodic musical tone signal, and also from time t2 to time t3. The signal portion up to is the second signal portion extracted as having a time width longer than the period of the lowest frequency in the audible frequency band for the signal corresponding to the portion other than the head of the aperiodic musical tone. It is.

In the previously proposed aperiodic musical tone signal generator,
Like the sounds of percussion instruments such as drums, cymbals, and various drums, the Ela is a non-periodic sound that corresponds to a sound that cannot be controlled in height, that is, an instrumental sound (aperiodic musical sound) that should not be included in a musical sound in the narrow sense. The periodic music T-f M signal is divided into the first signal part and the second /1411 part as described above, and each of them is sequentially stored in a digital (PI kJ-) Cl113 storage (read-only memory). I'll remember it :i
;The mode of reading the signal from the storage device is as described above, that is, the first readout of the non-periodic musical tone signal is:1:;
After the signal parts are read out in the storage order, the second 41'1 part stored in the read-only memory ROM is read out in the first readout mode, which is not in the storage order, and in the reverse storage order. A second readout mode such as a first readout mode and a second readout mode such as a non-readout mode are used to read out the signal in alternating 11, convert it into an analog signal, and then give it to the decimal circuit.
Fruit 4 in the diagram! i! By using the state expressed by the envelope as shown in the figure, like +J,
Even if a small amount of memory is used, it is possible to reproduce a non-periodic musical tone signal in good condition over a long period of time.

11- However, as shown in Figure 3, the actual sound of percussion instruments gradually decreases in amplitude on the time axis as one hour passes, and the high frequency components also decrease as time passes. Since the frequency spectrum distribution (spectrum envelope) is changing in such a manner as shown in FIG.
Reading is performed alternately using two readout modes: a first readout mode in which the reading is performed in accordance with the storage order, and a second readout mode in which the readout is performed in the reverse order of the storage order. When used as a signal with a state expressed by an envelope, parts with many high frequency components appear periodically, which may be perceived as noise-like sound, and the reproduced sound may be affected. This became a problem when it was necessary to increase fidelity. In FIG. 3, f is the frequency axis, t is the time axis, and A is the amplification axis.

(Means for Solving the Problems) The present invention provides a 12-1 signal portion corresponding to the beginning of an aperiodic musical tone and a portion other than the beginning of the aperiodic cold movement. A second signal portion extracted as having a time 111 longer than the period of the lowest frequency in the audible frequency band for the corresponding A number is sequentially stored in a storage device W and the storage device described above. After reading out the stored first part 1a in the order of storage, the memory 1
4i1! The second signal portion stored in i is read out using two readout modes: a first readout mode in which the second signal portion is not read out in accordance with the storage order, and a second readout mode in which the second signal portion is read out in the reverse order of the storage order. In the aperiodic musical tone (I'll) generator, which is provided with means for preventing alternate reading of the tones, it is used to store the first E part and the second A part. As a second signal part stored in the storage device in which the signal corresponds to a part other than the beginning of a non-calendar musical tone, An aperiodic method is used in which the short-time frequency spectrum distribution is approximately uniform regardless of the passage of time by transforming the fff part a) taken out as a Raku-a signal generator, and the above-described first and second signal band parts.
As a second signal part stored in the storage device used for storing the signal part, the period of the lowest frequency in the audible frequency band for the signal corresponding to the part other than the beginning of the aperiodic musical tone. A signal portion that is extracted as having a time width longer than use,
Also, memory S! The changes in the short-time frequency spectrum i-ram distribution on the time axis -1 in the signal portion sequentially read out from the time axis l2 on the original aperiodic musical tone signal on the time axis. An apparatus for generating an aperiodic musical tone signal is provided, which includes means for approximating the change mode of the short-term frequency spectrum distribution on the time axis -1 shown in FIG.

(Example) Hereinafter, the whiteness of the aperiodic musical tone signal generating device of the present invention will be explained in detail with reference to the accompanying drawings.

FIG. 4 is a block diagram of an embodiment of the non-periodic musical tone signal generating device of the present invention, in which RO
M is read-only memory,
This read-only memory ROM has a predetermined length of aperiodic data! The signal is encoded and stored as a digital signal. In the following explanation, it is assumed that the read-only memory ROM is a digital memory, but it goes without saying that the read-only memory used in implementing the present invention may be an analog memory. It is.

Now, in the non-periodic musical tone signal generator of the present invention, the read-only memories R and OM are used to record sounds with no perceptible pitch, such as the sounds of percussion instruments such as drums, cymbals, and various drums. In other words, an aperiodic musical tone signal corresponding to an instrument sound (aperiodic musical tone) that is not included in a musical tone in the narrow sense is in the first signal portion corresponding to the beginning of the aperiodic musical tone, etc. The signal portion extracted as having a period 11 longer than the period of the low frequency Jl that falls into the audible frequency band with respect to ri1 corresponding to the part other than the beginning of the aperiodic musical tone is defined as rj1 and a second signal portion transformed into one with a uniform wave number spectrum distribution, and are sequentially stored in a read-only memory ROM as digital signals.

The aperiodic musical tone signal stored in the read-only memory ROM as two signal portions as described above is read out after the first signal portion of the aperiodic musical tone signal is read out in the storage order. Read only memory R
The following describes two readout modes in which the second signal portion stored in the OM is read out according to the storage order, and a second readout mode in which the second signal portion is read out in the reverse order of the storage order. By alternately reading out the data, even if a memory with a small storage capacity is used, it is possible to successfully reproduce a non-periodic musical tone signal over a long period of time.

Here, in the aperiodic musical tone signal generating device of the present invention, a part of the aperiodic musical tone signal to be repeatedly used as described above will be explained with reference to FIGS. 1 and 3. This will be done with reference to FIG.

As mentioned above, Figure 1 shows the sounds of percussion instruments such as taiko drums, cymbals, various drums, etc., such as ffl, which have a pitch that cannot be felt, and instrumental sounds that are not included in the narrow sense of musical sounds (non-periodic sounds). 1 is a waveform diagram schematically showing an example of an aperiodic musical tone signal corresponding to a musical tone), and in FIG. 1, the horizontal axis is the time axis.

In Figure 1, I+, 5 ticks t1 to 11, v ticks t2
The part up to is the first signal part corresponding to the beginning of the aperiodic musical tone, and the part A from time t2 to time t3 in FIG. It should be made to correspond to other parts! ! An example of the original signal of the second signal part is shown (as the original signal of the second signal part to be created in response to λr for the part other than the beginning of the aperiodic musical tone, the original signal of the second signal part is (It may also be a signal in a portion away from the M band portion).

In addition, Fig. 3 is a diagram schematically showing how the short-term frequency spectrum distribution of aperiodic R & (M l) changes on the time axis. t
1. t2. t 3 etc. are the times t 1 . . . t 3 etc. shown in FIG. t2. This corresponds to t3.

(a) of FIG. 5 shows a portion Sa from time t1 to time t2 in the aperiodic musical tone signal shown in FIG.
That is, it is the first signal portion Sa corresponding to the beginning of the aperiodic musical tone, and (b) in FIG. 5 is the first signal portion Sa corresponding to the beginning of the aperiodic musical tone. This is the part sb up to time t3, that is, the part of the original signal sb of the second signal part to be created corresponding to the part other than the beginning of the aperiodic musical tone.

FIG. 5C shows a portion sb from time t2 to time t3 of the aperiodic musical tone signal shown in FIG. The original signal s of the second signal portion to be created corresponding to the portion of
This is the second signal Sbm obtained by transforming the signal Sbm into a state in which the short-time frequency spectrum distribution of the signal Sbm is almost flat regardless of the passage of time.

It is created corresponding to the portion sb from time t2 to time t3 in the aperiodic musical tone signal shown in FIG. 5(b), that is, the portion other than the beginning of the aperiodic musical tone. Should the second? From the original signal sb of the rt part, its short-time frequency spectrum distribution is transformed into one that is almost flattened regardless of the passage of time. Creating the signal Sbm of 2 and 4- is a frequency characteristic changing circuit (sound quality adjustment circuit).
This can be easily realized by applying well-known technical means.

Now, in FIG. 4, the DAC is a digital-analog converter that converts the digital signal read from the read-only memory ROM into an analog signal, and the LPF
is a low-pass filter, SAC is a signal processing circuit, MM is a monostable multivibrator, and 1 is an output terminal.

UDC is an up/down counter, and this up/down counter tJDc supplies an address signal to the read-only memory ROM. 1) Sl is a digit register provided to preset the initial values in the up/down counters T, JI-)C, and PS2. PS3 is a digitizer provided to preset up/down conversion points in the up/down counter UDC, and COMP is a comparator, F
F is a flip-flop, SW is a changeover switch, K is a key, GC is a gate circuit, and PG is a counting pulse generator.

In the following explanation, the up/down counter UDC is
Q of the high level state of the flip-flop FF mentioned above
When u small output is given to its terminal U, it operates as an up counter, and when the high level Qd output of the flip-flop FF mentioned above is given to its terminal d. It is supposed to operate as a down counter. Also, up/down counter U
The DC is conditioned by the flip-flop FF so that it is in one predetermined operating state (the predetermined operating state between the up counter state and the down counter state) in the initial state. However, in the following explanation, the up/down counter U D
C is said to operate as an up counter in its initial state.

Now, the above-mentioned read-only memories R and OM contain sounds such as drums, cymbals, various drums, and other percussion instruments that have a sense of pitch and the first sound, that is, are included in musical sounds in the narrow sense. The non-periodic music tIAη (non-periodic musical tone) and the corresponding non-periodic music 6 are stored in the form of a signal or digital signal, but in the read-only memory R mentioned above.
The manner in which the aperiodic musical tone signal is stored in the OM is as follows: the first signal portion Sa ((a) in FIG. 5) corresponding to the head of the aperiodic cold movement; From the original signal S h of the second signal part extracted as having a time width longer than the period of the lowest frequency in the audible frequency band for No. 4f1 corresponding to the part other than the beginning of the musical tone, The short-time frequency spectrum distribution is transformed into one that is instantly uniform regardless of the passage of time.
The second (tlsl
The number n is stored sequentially.

In the aperiodic musical tone generator (No. 11) of the present invention, the reading of the aperiodic musical tone signal stored in the read-only memory ROM is performed using the first signal in the aperiodic musical tone signal. After reading out the portions Sa in the storage order, the second signal portion Sbm stored in the read-only memory ROM is read out in the storage order; a first readout mode is such that the second signal portion Sbm is read out in the storage order; By using a special reading method in which reading is performed alternately using two reading modes, it is possible to produce good percussion instrument sounds over a long period of time even with a semiconductor memory having a small storage capacity. This has made it possible to provide a device for generating aperiodic musical tone signals that can be reproduced in a number of ways.

Continuation of storage signals from the above-mentioned read-only memory ROM by a special reading method can be easily carried out, for example, as follows. That is, reading of the storage signal from the read-only memory ROM is performed from the up-down counter UDC to the read-only memory R.
Since this is done according to the address signal supplied to OM, the above J
, 1 . /SphnI4・ may be determined in a predetermined manner, and in the illustrated embodiment, generation of the address signal as described above is performed using a plurality of digitizers, comparators, and flip-flops. It is made to be carried out by

In the embodiment shown in FIG. 4, when the key K is turned on, a counting pulse is sent to the up/down counter rJDC (I) via the circuit GC. , the terminal +1 of the up/down counter IJDC mentioned above is the flip-flop F.
Since the high 1 novel state is set by the Q u output of F, the up/down counter UDC is set to the numerical value Nls set in the digitizer PS1 from the initial state when the key K is turned on. is the initial value, performs the n+ number of up counter operations, and generates an address signal,
It also supplies the read-only memory ROM.

-23= The count value of the up/down counter UDC mentioned above is also given to the comparator GOMP, and the comparator C○MP calculates the value given to it via the fixed contact e and the movable contact V of the changeover switch SW. Numerical value N set in number counter PS2
e and the count value of the up-down counter UDC described above, and when the count value of the up-down counter UDC reaches the numerical value No. set in the digitizer PS2,
generate a coincidence pulse and apply it to the flip-flop FF;
The Qd output of the flip-flop FF is changed from a low level state to a high level state, and the operation mode of the up/down counter UDC, which was previously operating as an up counter, is changed so that it operates as a down counter. .

Further, as described above, when the Qd output of the flip-flop FF changes from a low level state to a high level state, the movable contact V of the changeover switch SW changes to the fixed contact e.
The comparator GOMP is switched from the fixed contact r side to the fixed contact r side, and the numerical value Nr set in the digitizer PS3 is given to the comparator GOMP via the fixed contact r and the movable contact V of the changeover switch SW. It will be done.

Then, the up/down counter UDC whose operating mode has been changed to operate as a down counter as described above.
11 value gradually decreases from the above-mentioned value No. and reaches the value N'', the comparator GOMP generates a coincidence pulse and applies it to the flip-flop FF,
I Tsubu 1z1? changes its Qd output from a high level state to a state of l':I = 1/bell, and the up/down counter 9 TJDC changes its operation mode from operating as a down counter to operating as an up counter. is changed.

Hereinafter, the above-mentioned up/down counter NT T)(';
Each time its count value reaches No from Nr, the operation mode is changed from up-counter operation to down-counter operation, and the up-down counter UD
Every time the count value of C reaches from No to Nr, the operation mode is changed from down counter operation to up counter operation, so that a III number operation is performed.

The signal read out by the address signal indicated by the numerical value Ne set in the digit register PS2 described above and the signal read out by the address signal indicated by the numerical value Nr set in the digit register PS3 described above are, for example, , the above-mentioned numerical value N e so that the signal is near the same polar peak value, respectively.
, Nr is set, even by switching between the first readout mode and the second readout mode for the second signal portion described above.

A reproduced signal with satisfactory waveform continuity at the repetition point of the second signal portion is obtained.

Further, the signal read out by the address signal indicated by the numerical value Ne set in the digit register PS2 described above and the signal read out by the address signal indicated by the numerical value Nr set in the digit register PS3 described above are , for example, the above-mentioned numerical value Ne, so that the amplitude of each signal is near zero,
Even if Nr is set and the polarity of the reproduced signal is inverted in response to switching between the first readout mode and the second readout mode for the second signal portion, the second A reproduced signal with satisfactory waveform continuity at the repetition points of the signal portion can be obtained.

An embodiment of the non-periodic musical tone signal generating device of the present invention shown in FIG. A non-periodic musical sound signal that corresponds to a sound of a percussion instrument such as a drum, whose pitch is difficult to perceive, that is, a non-periodic musical sound included in a musical sound in the narrow sense, is a non-periodic musical sound signal. The lowest signal in the audible frequency band for the first signal portion Sa ((a) in FIG. 5) corresponding to the beginning of the periodic musical tone and the signal corresponding to the portion other than the beginning of the aperiodic musical tone. In Fig. 5, the original signal sb of the second signal portion extracted as having a time width longer than the frequency period is transformed into one in which the frequency spectrum distribution is made uniform. The second signal Sbm shown in (c) is read out from the read-only memory ROM in which the second signal Sbm is sequentially stored as a digital E number. A first reading mode in which a first signal portion of the signal is read out in the storage order, and then a second signal portion stored in the read-only memory ROM is read out in the storage order, and a second signal portion is read in the opposite storage order. It is clear that the data is read out alternately and sequentially in two readout modes, such as the second readout mode, such as a second readout mode such as a second readout mode, and thereby, even with a semiconductor memory having a small storage capacity, the key K can be operated. This makes it possible to provide an aperiodic musical tone signal generating device that can satisfactorily reproduce long-term percussion instrument sounds over a period of time.

When the key K is operated in the aperiodic musical tone signal generator shown in FIG. 4, the pulses generated by the pulse generator PG are supplied to the up/down counter UDC via the gate circuit GC, as described above. At the same time, the pulse generated by the monostable multivibrator MM is applied to the up/down counter UDC by operating the key, so that the numerical value Ng set in the digitizer PS1 is preset to the up/down counter UDC.

28- Also, the pulses generated in the rB stable multivibrator MM by the operation of the above-mentioned key
1), the signal processing circuit SAC processes the signal supplied from the low-pass filter [, I) F into a predetermined envelope (e.g., a monotonically decreasing envelope). In addition, the manner of change in the frequency spectrum distribution on the time axis in the signal is different from the manner in which the frequency spectrum distribution changes on the time axis in an actual percussion instrument. As exemplified in FIG.

When the second signal portion stored in the read-only memory ROM is the second signal Sbm as shown in FIG. 5(c), the signal processing circuit SAC described above is executed. 1 For example, in addition to the DIKE circuit configured by combining a delay circuit and a voltage control amplifier, there is also a frequency characteristic changer that operates to gradually reduce high-frequency signal components for second signals that are sequentially read out. A device that includes a circuit can be used.

As a result, the changes in the amplitude level and short-time frequency spectrum distribution on the time axis of the signal after passing through the signal processing circuit SAC described above are as follows: This can be approximated to the manner in which the amplitude level and short-term frequency spectrum distribution shown on the axis change over time.

The signal processing circuit SA is configured to gradually reduce the high-frequency signal components of the second signals sequentially and alternately read from the read-only memory ROM on the time axis.
As the frequency characteristic changing circuit to be provided in C, for example, a digital filter can be used, and the transfer characteristic of the digital filter can be changed at one or more arbitrarily specified points on the time axis. According to
It is possible to easily generate aperiodic musical tones that are audibly good.

FIG. 6 is a signal waveform diagram of each part of the aperiodic musical tone signal generator shown in FIG. 4, and (Il) in FIG. I)) is the pulse generated by the monostable multivibrator MM, (C) in Figure 6
is the Q u output of the trip-flop fp, and (d
) indicates the coincidence output of the comparator (', OM l', and (,) in FIG. 6 indicates the signal sent to output terminal 1.

In the embodiment of the aperiodic musical tone tn >+ generation device of the present invention described above, the signal readout for the second No. 48 portion in the first readout mode and the second readout mode is performed in the second readout mode. July 16 specific part in the signal part (W several books PS2
The numerical value NO set in and the numerical value N set in the digit register PS3!・Toyo-)) However, in implementing the present invention,

The signal readout by the first readout Ml method and the second readout method for the second signal portion is performed on a part of the second ff1 portion, or the second signal portion The signals are read out in the first readout mode and the second readout mode for a part of the 31-second signal portion in such a manner that the signal readout range is random. As described above, the reading of the signal in the first readout mode and the second readout mode for the second signal portion causes the signal to be read out for a part of the second signal portion. If the readout range is random, the state of the reproduced sound becomes more aperiodic, and the reproduced sound has a better audible quality.

FIG. 7 shows that the aperiodic musical tone signal generating device of the present invention is implemented so that the range of signal readout for the second signal portion by the first readout mode and the second readout mode is random. FIG. 4 is a block diagram showing an example configuration of a read range designation circuit ASG portion when the reading range is specified.

In FIG. 7, the read-only memory ROM a includes:
Numerical values to be given one after another to the comparator COMP are stored, but the stored numerical values one after another are given to the comparator GO.
When given to MP, the readout range 32- of the signal for the second signal portion is random, and the continuity of the waveform at the repetition point of the second signal portion is perfect. Numerals such as this are being made.

The odor of the example configuration shown in FIG. 7'c.

CT is a counter, P G a is a pulse source, and key K
is operated, as in the case of the embodiment of the aperiodic musical tone signal generator shown in FIG. At the same time, the above-mentioned counter CT is set to be in a state where it is possible to perform n1 number of operations. When the first address of the read-only memory 11 Il Ma is designated thereby, the value stored at the first assigned address (i.e., the first address stored in the read-only memory ROM) is The address at the end of the signal part and the corresponding number are read-only.
- Transferred from memory ROM a to comparator COMT'.

The 11 values of up/down counter U I) C are
From the read-only memory ROMa mentioned above to the comparator COM
-If the value given to P, the comparator GOM
Another pulse is output from P, and the counter CT is incremented by 1 ink by the coincidence pulse of 111 and (, and the next address is specified in the read-only memory R, OM port.

Read only memory ROM a to J:? This is because the addresses after the specified address of iJru(tri) are stored with numerical values that can randomly specify the read range of the second f4 part stored in the read-only memory ROM. As the counter Cl' is incremented each time one pulse is output from the comparator COM I), the read only memory I? Numerical values that can randomly specify the readout range of the second signal portion stored in OM are output to the comparator COMP one after another.

Therefore, by using the read range designation circuit ASG as illustrated in FIG. The reading range of the signal is random for a part of the signal portion of the signal.
The state of the reproduced sound becomes more non-periodic, and the reproduced sound has a better audible state.

(Effects) As is clear from the detailed explanation above, the aperiodic musical tone signal generation device of the present invention has a first signal portion corresponding to the head of the aperiodic musical tone, and and a second signal portion extracted as having a time width longer than the period of the lowest frequency in the audible frequency band for a signal corresponding to a portion other than the head of the aperiodic musical tone. a storage device; and a first reading mode in which, after reading the first signal portions stored in the storage device in the storage order, the second signal portions stored in the storage device are read out in the storage order. , and a second readout mode in which the reading is performed in the opposite order to the storage order. In this case, the beginning of an aperiodic musical tone is detected in the second signal portion stored in the storage device 35 used for storing the first signal portion and the second signal portion. The short-time frequency spectrum distribution of the signal part extracted as having a time width longer than the period of the highest OL frequency in the audible frequency band for the 1M signal corresponding to the part other than the A device for generating an aperiodic musical tone signal, which is used by transforming into a state in which it is substantially unified regardless of the passage of time,
and a second signal stored in the storage device used to store the first signal part and the second (PI part).
As the signal part of the signal corresponding to the part other than the beginning of the aperiodic musical tone, the signal part extracted as having a period of time [11] longer than the period of the low frequency in the audible frequency band is defined as , the short-time frequency spectrum distribution is transformed into a state in which it is uniformized regardless of the passage of time over the time 111, and is used, and the time in the signal portion that is sequentially read out from the storage device on the time axis is used. To approximate the change mode of the short-time frequency spectrum distribution on the axis to the change mode on the time axis of the short-time frequency spectrum 1 to Lamb distribution that the original aperiodic musical tone No. 41 shows on the time axis. Since the present invention is a non-periodic musical tone signal generating device having a means for generating a signal, the present invention uses a semiconductor memory with a small storage capacity, similar to the previously proposed non-periodic musical tone signal generating device. Of course, it is possible to easily provide a non-periodic musical tone signal generating device that can satisfactorily reproduce long-duration percussion instrument sounds. There is no noise that can occur in aperiodic musical tone signals generated by aperiodic musical tone signal generators, and there is no noise that can occur in aperiodic musical tone signals generated by aperiodic musical tone signal generators that have been proposed. This makes it possible to easily generate percussion instrument sounds that sound more natural than conventional musical sound signals.

[Brief explanation of drawings]

Figures 1, 2, and 5 are signal waveform diagrams;
The figure shows the characteristic number of times, Figure 4 is a block diagram of an embodiment of the non-periodic musical tone signal generating device of the present invention, Figure 6 is a waveform diagram of signals at various parts of the device, and Figure 7 is the specification of the continuous range. FIG. 2 is a block diagram of an example circuit. ROM, ROMa...Read-only memory, ri A
C...DA converter, LPF...low pass filter,
1...Output terminal, UDC...up/down counter, Psi~PS3...digitizer, COM F+...comparator, FF...flip-flop, SW...changeover switch, K...・Key, GC...Gate circuit, PG
* PG a...Pulse generator, CT...Counter, ASG...Reading range designation circuit, 39-', t2 1→ 1 Redundancy 1t\1 Tail 3 Kasumi Continuation of 1st page 0 shots Akiyoshi 1) Hiroshi Yokohama City Kanazuki Company

Claims (1)

[Claims] 1. The lowest frequency in the audible frequency band for the first signal portion corresponding to the head of the aperiodic musical tone and the signal corresponding to the portion other than the head of the aperiodic musical tone. A storage device which sequentially stores the second (No. 6 part) taken out as having a period 111 longer than the cycle of A first readout mode in which the second signal portion stored in the storage device is read out in the order in which the signal is stored in the storage device, and a second readout mode in which the second signal portion is read out in the reverse order of the storage order. In an aperiodic musical tone signal generating device comprising means for alternately reading out data in a reading mode, the device is used to store the first signal portion and the second signal portion. The second signal portion stored in the storage device is J: Una, which has a time width longer than the period of the reddest frequency in the audible frequency band for a signal corresponding to a portion other than the head of an aperiodic musical tone. Generating an aperiodic musical tone signal by transforming a signal portion extracted as a signal into one in which the short-term frequency spectrum distribution is almost flat regardless of the passage of time. The device 2 detects the lowest frequency in the audible frequency band for the first (t'1 part) corresponding to the beginning of the aperiodic musical tone and the signal corresponding to the part other than the beginning of the aperiodic musical tone. a storage device for sequentially storing a portion of the second signal extracted as having a time period better than the period, and a first signal portion stored in the storage device; #I reading out the second signal portion stored in the storage device according to the storage order after reading out the second signal portion in the storage order;
11! ! The two readout modes, the first readout mode and the second sequential readout mode, which read out in the opposite order of storage, are interchanged.
In the aperiodic musical tone signal generation device, the signal portion 111111 and the second signal portion are stored in the storage device used for storing the signal portion 111111 and the second signal portion. As the second signal part, a signal part extracted as having a time width longer than the period of the lowest frequency in the audible frequency band is used as the second signal part of the signal corresponding to the part other than the beginning of the aperiodic musical tone. It is used by transforming the short-time frequency spectrum distribution into a state in which it is almost flat regardless of the passage of time over the time span, and also uses the time axis in the signal portion that is sequentially read out from the storage device on the time axis. Short-time frequency spectrum 1 above
A non-periodic signal that includes a means for making the variation of the Hellam distribution approximate the variation on the time axis of the short-term frequency spectrum distribution shown on the time axis by the original aperiodic musical tone signal. A musical tone signal generating device 3 calculates the changes in the short-time frequency spectrum distribution on the time axis in the signal portions sequentially read out from the storage device on the time axis. As a means for approximating the changes in the short-time frequency spectrum distribution on the time axis shown in (2) and (2), there is a method that can change the short-time frequency spectrum distribution from an arbitrarily specified time on the time axis (2). The aperiodic musical tone signal generating device 4 according to claim 2 is used, and the short-time frequency spectrum distribution on the time axis in the signal portions sequentially read out from the storage device on the time axis. As a means of making the change mode approximate the change mode on the time axis of the short-time frequency spectrum distribution that the original aperiodic musical tone signal shows on the time axis, A device for generating an aperiodic musical tone signal using a device that can change the frequency spectrum distribution over a short period of time several times apart.