JPS6055398A - Waveform formation for electronic musical instrument - Google Patents

Abstract

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

Description

DETAILED DESCRIPTION OF THE INVENTION Technical Field The present invention relates to a waveform forming method for an electronic musical instrument, and in particular a method for preparing a waveform memory with two types of waveforms each having a plurality of cycles and having different characteristics, and interpolating and synthesizing the waveforms read from both memories. This paper relates to a waveform forming method for an electronic musical instrument.

Conventional technology The entire waveform from the start to the end of the sound, or the entire waveform at the rising edge and part of the subsequent waveform, is stored in a waveform memory, and if the former is stored, the entire waveform is read out in order to create a high-quality musical sound waveform. Recently, when a signal is generated and the latter is stored, a high-quality musical waveform signal is generated by reading out the waveform at the rising edge and then repeatedly reading out part of the waveform after that. . Although this method of storing multi-cycle continuous waveforms in the waveform memory in advance can provide high-quality musical waveform signals, it requires a huge amount of memory capacity, so it is possible to It was unsuitable for achieving significant tonal changes. In other words, the simplest solution would be to prepare in advance a large number of different waveform memories that correspond to all types of key touches or tone-level timbre change parameters, but this would require too much memory capacity. It becomes huge and impractical.

It is true. Therefore, one method is to prepare two types of continuous waveforms in the waveform memory, for example, in the case of touch response control, a continuous waveform corresponding to the strongest touch and a continuous waveform corresponding to the weakest touch. By reading out and interpolating both waveforms according to the timbre change parameter (touch intensity), the timbre change parameter (touch intensity) is determined.
It is conceivable to obtain a waveform corresponding to , but in reality, if the phases of both waveforms to be interpolated do not match, the interpolation becomes meaningless. The two types of waveforms to be prepared in the waveform memory are copies of the actual sound waveforms played, so the phases of both waveforms are different, and even if the initial phases can be matched, a large phase difference will occur after a few seconds. arise. Therefore,
Simple interpolation is not suitable for methods that attempt to obtain high-quality musical waveform signals by storing multi-cycle continuous waveforms in memory and reading them out, and it is necessary to realize a variety of timbre changes with a small-scale configuration. This has traditionally been difficult.

Purpose of the Invention Therefore, the purpose of the present invention is to make it possible to actually perform interpolation without any inconvenience by using two types of multi-period waveforms obtained based on natural musical instrument sounds, and to achieve this on a relatively small scale and at low cost. The objective is to realize high quality and diverse timbre changes with a simple configuration.

Summary of the Invention The waveform forming method according to the present invention includes (1) preparing two types of original waveforms each consisting of a plurality of cycles produced by the same type of natural musical instrument with mutually different timbre and volume characteristics; (2) preparation. By manipulating the phase of the original waveform, the phase is corrected in the direction of reducing the mutual phase shift so that the mutual phase relationship does not deviate too much, and as a result, the first and second waveforms consisting of multiple periods are created. (3) Preliminarily storing the entire waveform or the rising part and a part of the waveform after the first and second waveforms obtained in (2) above in the first and second waveform memories, respectively. - (4) the first pitch according to the specified pitch;
and obtaining a desired waveform signal according to the timbre change parameter by reading the waveforms from the second waveform memory and interpolating and synthesizing the read outputs of both memories at a ratio according to the timbre change parameter. It is characterized by becoming. By operating the phases of two types of waveforms obtained based on natural musical instrument sounds as described above and preventing the phase shift between them as much as possible, interpolation synthesis using both waveforms can be performed without any inconvenience.

According to the present invention, in order to further simplify the storage contents of the waveform memory, the difference at each sample point of the two original waveforms whose phases have been corrected as described above is calculated to obtain a differential waveform between the two waveforms, and the phase It is proposed to store one of the modified original waveforms in a first waveform memory and the difference waveform in a second waveform memory. Since each sample point data of the differential waveform can be expressed with a smaller number of bits than normal waveform sample point amplitude data, the capacity of the second waveform memory can be reduced.

The present invention can be applied to touch response control that controls the timbre and level according to the touch of a key, key scaling control that controls the timbre and level according to the pitch or range of the pressed key, and other timbre change controls. . Therefore, as the timbre change parameter, the intensity of the key touch, the pitch of the pressed key, its range, or other factors that promote the timbre change are used.

Embodiment FIG. 1 is an electrical block diagram showing an example of an electronic musical instrument related to the implementation of the present invention, in which a keyboard 10 is used as a means for specifying the pitch of a musical tone to be generated. A touch applied to a key is detected by a touch detection device 11, and this touch detection data is used as a timbre change parameter to generate a musical waveform signal having timbre and level characteristics depending on the intensity of the touch. The first waveform memory 12 stores in advance all waveforms of the following characteristics (temporarily referred to as the "first waveform") from the rising edge of sound generation to the end thereof.

The second waveform memory 16 stores information regarding a waveform (temporarily referred to as a second waveform) having characteristics different from the first waveform.
The entire waveform from the start of the sound to the end is stored in advance. In this example, the first waveform has a characteristic corresponding to the strongest key touch.

The second waveform is a waveform with characteristics corresponding to the weakest key touch.

An address data generation circuit 14 provided between the keyboard 10 and the waveform memo IJ 12 , 13 generates the waveform memo IJ 12 , 1 from the beginning to the end of sound generation according to the pitch specified on the keyboard 10 . This is a reading means for reading out all waveforms respectively. Waveform memo IJ 12 according to address data generated by address data generation circuit 14
, 13 respectively read out the first and second waveform signals,
It is input to the interpolation means 15. The interpolation means 15 interpolates both waveform memories 1 at a ratio according to the touch detection data (i.e. timbre change parameter) given from the touch detection device 11.
2.16 output signals are synthesized to obtain a waveform signal having characteristics corresponding to the strength of the key touch.

In the interpolation means 15, the interpolation coefficient memories 16 and 17 store in advance 2 in the first and second interpolation coefficients Kl, which have opposite characteristics with respect to the touch intensity, and perform the interpolation according to the touch detection data. Read out 1°K2 as a coefficient. The first interpolation coefficient of 1 means that the level increases as the touch intensity increases, and the level of the waveform signal corresponding to the strongest touch read from the first waveform memory 12 is determined by the first interpolation coefficient in the multiplier 18. It is controlled according to a coefficient of 1. The second interpolation coefficient of 2 means that the level increases as the touch strength becomes weaker.
The level of the waveform signal corresponding to the weakest touch read out from the second waveform memory 16 is determined by the multiplier 19.
is controlled according to the interpolation coefficient of 2. Multiplier 18.1
The outputs of 9 are added by an adder 20 to obtain an interpolated and synthesized waveform signal. This waveform signal is converted into an analog signal by a digital-to-analog converter 21, and then converted to an analog signal by a sound system 2.
given to 2.

The first and second waveforms to be stored in the first and second waveform memos IJ 12 and 13 are created by the following preprocessing.

Process 1: A predetermined natural musical instrument (for example, a piano) is played with a weak touch, and an original waveform A of multiple cycles from the start to the end of sound generation as shown in FIG. 2(a)l is obtained. When playing the same natural instrument with a strong touch, the original waveform B as shown in Figure 2 (FJ) is obtained.
get. Note that both original waveforms A and B have the same pitch.

Process 2... The phases of the original waveforms A and B prepared as described above are manipulated to correct the phases in a direction that reduces the mutual phase shift so that the mutual phase relationship does not shift too much. This phase operation is
- For example, as shown in the following processes 2a to 2C, one of the original waveforms B is filtered to obtain a waveform that approximates the other original waveform A.

Processing 2a... All waveform sections of original waveforms A and B are divided into multiple frames (time frames)
The spectrum analysis of both waveforms is performed for each frame. This frame division is not limited to equal time intervals, but is set at appropriate intervals depending on the characteristics of the waveform change. In the example shown in the figure, it is divided into 7 frames from 0 to 6 frames. An example of spectrum analysis results for one frame consisting of original waveform A
is as shown in Figure 3(a), and the original waveform B is as shown in Figure 3(1).
It seems like no.

Processing 2b... The deviation between both spectra in the same frame analyzed in Processing 2a is calculated for each frame. For example, Figure 3 Ca),
The spectral deviation of (1)) is as shown in (C).

Processing 2c...Filter characteristic parameters for each frame are determined based on the spectral deviation for each frame determined in Processing 2b, and filter operations are performed for each frame on the original waveform B corresponding to the strong touch according to the filter characteristic parameters. administer. Through this filter operation, a waveform that approximates the original waveform A corresponding to a weak touch can be obtained.

After the above process 2c, the original waveform B corresponding to the strong touch that was the target of the filter operation is stored in the first waveform memory 12, and a waveform that approximates the original waveform A corresponding to the weak touch obtained by the filter operation is stored. is stored in the second waveform λ' memory 16. In this way, the waveform stored in the second waveform memory 16 approximates the original waveform A, but the original waveform B has been filtered.

Therefore, its phase does not deviate too much from the original waveform B. Therefore, both waveform memo I
It is possible to prevent the phases of the waveforms stored in J 12 and J 13 from shifting too much.

As another example of the phase manipulation in the process 2, a method shown in the following process 2' may be used.

Process 2'... By shifting the phase of one or both of the original waveforms A and H by an appropriate amount for each predetermined phase interval, both waveforms A and H in each phase interval are
The phase is corrected in the direction in which the phases of B match. thus,
Original waveform A with phase correction.

Store B in waveform memory 1) 13.12, respectively. Such phase manipulation can also prevent the phases of the waveforms stored in both waveform memories 12 and 13 from shifting too much.

As described above, the waveform memories 1) 12 and 13 store a waveform corresponding to a strong touch and a waveform corresponding to a weak touch, whose phases are manipulated so that their phases do not deviate too much. Therefore, in the interpolation means 15, both waveforms 12.13
When interpolating and synthesizing the outputs of , it is possible to synthesize the outputs without any inconvenience (without generating beats or the like caused by phase shifts).

FIG. 4 is an electrical block diagram of an electronic musical instrument according to another embodiment of the present invention, showing only the changes from FIG. 1.
Alternatively, the waveform corresponding to the weakening notch of the phase-corrected original waveform is stored as shown in 2'. In this embodiment, the following process 3 is performed after the above-mentioned processes 1 and 2 (2a to 2C, 2').
or added.

Processing 3: The difference between one and the other of the phase-corrected original waveforms obtained as a result of the phase manipulation in Processing 2 is determined for each sample point, and as a result, the difference between the two phase-corrected waveforms is calculated. Get the waveform.

In this way, the difference waveform obtained in process 3 is stored in the second waveform memory 2.
Remember it in 4. The interpolation means 15A includes an interpolation coefficient memory 25 that reads out an interpolation coefficient K according to touch detection data, a multiplier 26 that multiplies this interpolation coefficient by the difference waveform data read out from the second waveform memory 24, and this multiplier. The output of the device 26 is read out from the first waveform memory 23 (
(waveform corresponding to weak touch). The coefficient memory 25 reads "]" as a coefficient when the key touch is the strongest, reads "0" as K when the key touch is the weakest, and selects a coefficient that satisfies the condition O<K<1 depending on the touch strength during that time. K is read out according to a predetermined interpolation function. In this way, the addition ratio of the differential waveform data to the weak touch compatible waveform read from the memory 23 is controlled according to the key touch strength, and a waveform with characteristics according to the touch strength is obtained.

By the way, the difference waveform stored in the memory 24 is the difference in amplitude value for each → Kanpur point of the waveform corresponding to strong touch and the waveform corresponding to weak touch, so it is a spiky waveform with many harmonics. be. When this thorny difference waveform is added, even at a small level, to the weak touch compatible waveform, does it feel like the summed and combined waveform suddenly changes from the weak touch compatible waveform read out from the memory 26? There is a risk that the waveform will be different from the sound of an actual natural musical instrument performance. Therefore, the interpolation means 15A is changed as shown in FIG. 5, and a digital filter (low-pass filter) 28 is provided on the output side of the second waveform memory 24. It is preferable to read the corresponding filter characteristic parameters and control the filter 28 accordingly. In this filter control, the weaker the key touch, the more rounded the difference waveform is output from the filter 28, and the stronger the touch, the less rounded the difference waveform is output from the filter 28, which is closer to the original difference waveform output from the waveform memory 24. Make it output. Then, when the touch is the strongest, the output waveform of the waveform memory 24 is output from the filter 28 without making any changes. With this kind of control, the waveform corresponding to the weakest touch (output of the memory 26) is generated before the touch is relatively weak.
The difference waveform added to can be made smooth with less harmonic content, and the above-mentioned disadvantages can be eliminated.

Note that the first waveform memory 26 may store the strongest touch-compatible waveform, and the adder 27 may be used as a subtracter.

In each of the above embodiments, the waveform memo IJ 12, 13
.

In 23 and 24, the entire waveform from the start to the end of sound generation is stored, but the present invention is not limited to this, and the waveform at the rising edge and part of the waveform after it may be stored. In that case, the address data generation circuit 14 reads out the waveform at the rising edge, and then repeatedly reads out part of the subsequent waveform (this is also a multi-cycle waveform), so that Make sure to read out all waveforms.

The amplitude envelope of the waveform signal read out repeatedly is
It is applied by an appropriate envelope applying means.

In addition, the waveform memo IJ 12, 1, S, 23 does not store the musical waveform amplitude sample data as is, but stores the difference data between adjacent sample amplitude values, and when reading, this difference data is cumulatively stored. The original amplitude sample data may be obtained by adding or subtracting the .

In addition, when key scaling control of timbre is performed using the pitch or range of the pressed key as the timbre change parameter, the key touch strength or touch detection data in the description of each of the above embodiments can be read as the pitch or range of the pressed key. It can be implemented in exactly the same way.

Effects of the Invention As described above, according to this invention, two types of waveforms consisting of multiple periods obtained based on natural musical instrument sounds are phase-manipulated to prevent the phase of the two from shifting too much, and then Since the respective waveforms are stored in the waveform memory and the outputs of the waveform memory are interpolated and synthesized, it becomes practically possible to interpolate and synthesize two types of waveforms that approximate natural instrument sounds whose waveforms change over time. As a result, it becomes possible to realize high-quality and diverse timbre changes with a relatively small-scale and low-cost configuration. Furthermore, by storing the differential waveform in one waveform memory, the memory configuration can be further simplified.

[Brief explanation of the drawing]

FIG. 1 is an electrical block diagram of an electronic musical instrument according to an embodiment of the present invention, and FIG. Figures illustrating waveforms, Figure 3 (a) and (1)) are the second
Figures (a) and (b) show an example of the spectrum in one frame consisting of the waveform of J. Figure 3 (C,) shows (a
) and (!]-; FIG. 4 is an electrical block diagram showing an electronic musical instrument according to another embodiment of the present invention with respect to changes from FIG. 1; and FIG. Fourth
FIG. 6 is a block diagram showing a modification example of the interpolation means shown in the figure. 10 keyboard, 11...touch detection device, 12,1. S. 23.24・Waveform memory, 14 A・Res data generation circuit, 15, 15A・Interpolation means, 17, 16゜25
Interpolation coefficient memory. Applicant Nippon Gakki Mfg. Co., Ltd. Agent Yoshihito Iizuka''

Claims (1)

[Claims] 1. A pitch specifying means for specifying the pitch of a musical sound to be generated, and with respect to the first waveform, the entire waveform from the start of sound generation to the end, or the waveform of the rise. Regarding the second waveform, a first waveform memorizer stores a part of the subsequent waveform, and a second waveform memorizer stores the entire waveform from the start of the sound generation to the end, or the waveform of the rise and a part of the subsequent waveform. a waveform memory, a reading means for respectively reading out all waveforms from the start to the end of sound generation from the first and second waveform memories in accordance with the pitch specified by the pitch specifying means, A method for forming a waveform in an electronic musical instrument, comprising interpolation means for synthesizing the outputs of both waveform memories at a corresponding ratio, the method comprising a plurality of periods produced by the same type of natural musical instrument with mutually different timbre and volume characteristics. preparing two types of original waveforms; manipulating the phase of the prepared original waveforms to correct the phases in a direction that reduces mutual phase shift; as a result, the first and second waveforms each having a plurality of periods; and converting the first and second waveforms obtained above into the first waveforms as described above.
and storing in advance in a second waveform memory, and by combining the outputs of the first and second waveform memories read by the reading means by the interpolation means, the first and second waveform memories are stored in advance. A method for forming a waveform in an electronic musical instrument, comprising: forming a waveform signal by synthesizing a second waveform at a ratio according to the timbre change parameter. 2. The phase operation is to obtain a waveform that approximates the other original waveform by performing a filter operation on one of the original waveforms, and one of the original waveforms subjected to the filter operation is used as the first waveform. 2. The waveform forming method for an electronic musical instrument according to claim 1, wherein the second waveform is a waveform obtained by filtering. 3. The above-mentioned phase operation is to correct the phase in a direction in which the phases of both waveforms match in each phase interval by shifting the phase of one or both of the original waveforms by an appropriate amount in each predetermined phase interval, and in this way, the mutual phase A waveform forming method for an electronic musical instrument according to claim 1, wherein the first and second waveforms are two waveforms whose relationship has been corrected. 4. The waveform forming method for an electronic musical instrument according to claim 1, wherein the timbre change parameter indicates the intensity of a touch applied to a key pressed on a keyboard serving as the pitch specifying means. 5. The waveform forming method for an electronic musical instrument according to claim 1, wherein the timbre change parameter indicates the pitch specified by the pitch specifying means or its range. 6. A pitch specifying means for specifying the pitch of a musical tone to be generated, and regarding the first waveform, the entire waveform from the start of sound generation to the end, or the waveform of the rise and a part of the waveform after that. a first waveform memory that stores the second waveform, and a second waveform memory that stores the entire waveform from the rising edge of sound generation to the end, or the waveform of the rising edge and part of the subsequent waveform; Reading out all waveforms from the start to the end of sound generation from the first and second waveform memories, respectively, in accordance with the pitch designated by the pitch designation means. and interpolation means for synthesizing the output signal of the second waveform memory with the output signal of the first waveform memory at a ratio according to a timbre change parameter. , prepare two types of original waveforms consisting of multiple cycles produced by the same type of natural instrument with different timbre and volume characteristics, and reduce the mutual phase shift by manipulating the phase of the prepared original waveforms. The difference between one and the other of the phase-corrected original waveforms is determined for each sample point, and as a result, a difference waveform between both waveforms is obtained. One of the phase-corrected original waveforms is The first waveform is stored in advance in the first waveform memory as described above, and the difference waveform obtained above is stored in advance as the second waveform in the second waveform memory as described above. The outputs of the first and second waveform memories read out by the readout means are combined by the interpolation means, thereby changing the timbre of the second waveform to the first waveform. 1. A waveform forming method for an electronic musical instrument, comprising: forming a waveform signal synthesized at a ratio according to parameters. 7. The interpolation means includes means for generating a level coefficient according to the timbre change parameter, and level control of the output signal of the second waveform memory according to the level coefficient to output the first waveform memory. 7. The waveform forming method for an electronic musical instrument according to claim 6, further comprising calculation means for adding or subtracting from a signal. 8. The interpolation means includes means for generating a level coefficient according to the timbre change parameter, means for generating a filter characteristic parameter according to the timbre change parameter, and a means for generating a filter characteristic parameter according to the timbre change parameter; A digital filter that controls characteristics according to characteristic parameters,
means for controlling the level of the output signal of the second waveform memory controlled by the digital filter according to the coefficient; and the second waveform memory controlled by the digital filter; 7. The waveform forming method for an electronic musical instrument according to claim 6, further comprising arithmetic means for adding or subtracting the output signal of the waveform memory and the output signal of the first waveform memory.