JP2526124Y2 - Envelope extraction device - Google Patents

Description

[Detailed description of the invention] [Technical field of the invention] The invention relates to an envelope extracting device that extracts an envelope waveform from sampled sound waveform data.

[Background] The ability to synthesize natural musical instrument sounds is an important issue for electronic musical instruments, but to date, there is a considerable gap between electronically synthesized sounds and natural musical instrument sounds. ing. Electronically synthesizable sounds are closely related to the limitations of the processing capabilities of current musical sound processors. The music processing device
It does not have the ability to execute complex performances in real time, and can perform only simple, low-degree processing. As a result, most of the features such as complex tone changes seen in natural musical instrument sounds will be lost.

For example, in the case of a device that synthesizes a musical tone by multiplying data for one waveform of the fundamental frequency of the musical tone by envelope data, it is not possible to generate an envelope waveform that changes in a complicated manner. Sectionally monotonous waveform (eg exponential curve)
And changing the coordinates of the end of the interval is the only degree of freedom.

The sound-making method that is currently performed under the constraints of hardware and software accepts the constraints,
Within the few degrees of freedom available, trial and error,
Another approach is to artificially create sound. This goes in a completely different direction from the original approach to natural instrument sounds.

On the other hand, there is also known an apparatus which samples a waveform of a natural instrument sound by PCM and reads out the sampled waveform data in real time to generate a sound, as represented by a sampling instrument. This type of device can faithfully reproduce sampled sounds. However, since the musical sound information is in the form of a completed waveform, it is not suitable for arithmetic processing, and the characteristics of natural musical instrument sounds that vary in a complex manner depending on the pitch and playing style are created from the waveform data of one sampled natural musical instrument sound. Can not. in short,
It can be said that sampling musical instruments and the like basically abandon the idea of synthesizing sounds electronically.

In order to obtain a sound that sounds like a natural instrument by synthesis, the idea of analyzing the data, extracting the characteristics of the sound, and obtaining a synthesized sound using the extracted characteristic parameters is effective. It was devised with the recognition that it is.

[Purpose of the Invention] The purpose of the invention is to provide an envelope extraction device that extracts envelope data that preserves the characteristics of the original sound from sampled waveform data.

According to the present invention, in an envelope extracting apparatus for extracting envelope data from digital waveform data of a sound digitally sampled at a predetermined sampling frequency, a fundamental frequency representing a fundamental frequency of the sampled sound Basic frequency data inputting means for inputting value data, predetermined section length determining means for determining a data section of a predetermined cycle related to the digital waveform data from the sampling frequency value and the basic frequency value data, and the digital waveform data And the equal division means for equally dividing the entire waveform section into small sections having the length of the data section of the predetermined period, and the waveforms contained in the small sections for each of the small sections equally divided by the equal division means. Maximum / minimum value determining means for obtaining the maximum and minimum values of the data,
A difference calculating means for calculating a difference between a maximum value and a minimum value of each small section; and an envelope value determining means for determining an envelope value of each small section according to the difference calculated for each small section. Is provided.

In this configuration, the maximum value and the minimum value obtained for each small section respectively represent the maximum, that is, the positive peak and the minimum, that is, the negative peak, of the waveform of the small section. According to the above configuration, since the difference between the maximum value and the minimum value is obtained for each small section, when the waveform of the small section is asymmetric, that is, when the magnitudes of the positive peak and the negative peak are not equal. Even in this case, it is clear that the asymmetry can be compensated by calculating the difference between the maximum and the minimum. Further, according to the invention, all the waveform sections of the digital waveform data are equally divided into small sections of equal length by the equal dividing means. This guarantees sections of equal length as extraction sections for extracting the envelope. Further, according to this invention, the length of a small section, which is a unit section for envelope data extraction, is set based on the value of the sampling frequency and the basic frequency value data from the input means. For example, a data section having a length corresponding to the basic cycle of a sampled sound can be used as a small section. Such a data section of the musical basic cycle has a sampling frequency of Fs, and a basic frequency of the input sampled sound. The value can be obtained immediately by calculating Fs / Freq with Freq as the value.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

FIG. 1 shows the overall configuration of this embodiment. The musical tone data memory 1 stores digital waveform data (hereinafter referred to as musical tone data) of a sound sampled at a predetermined sampling frequency (Fs, which will be described later). Then, an envelope value is extracted in units of an extraction section of a predetermined cycle. The extracted envelope value is extracted. The extracted envelope data is stored in the extracted envelope memory 3. In terms of hardware, the envelope extraction circuit 2 includes an input / output device such as a keyboard and a display, a program and a working memory, a processor device, and the like.

FIG. 2 shows a configuration of a main part of the envelope extraction circuit 2. The extraction section setting scanning unit 21 receives the information from the input device, determines sections for extracting the envelope values one by one, and when extracting the envelope data, every time the envelope value of one section is determined, the next section is selected. Then, the entire section of the sampled tone data is scanned. In this example, sampling frequency information of musical sound data and musical sound (basic) frequency information are given as input information. Therefore, a section corresponding to a cycle of a sampled musical tone is selected as a section (extraction section) used to extract one envelope value.

The in-section musical sound data reading unit 22 sequentially reads out the musical sound data included in the current section given by the extraction section setting scanning unit 21,
The data is read from the tone data memory 1.

The maximum / minimum extraction unit 23 extracts the maximum value and the minimum value of the musical tone data in the section provided by the reading unit 22. When the completion of the current scanning section is notified from the extraction section setting scanning unit 21, the extracted maximum The difference between the value and the minimum value is calculated, and this difference is written as the envelope value of the section of interest in the memory position assigned to the current scanning section number indicated by the extraction section setting scanning section 21.

FIG. 3 is a flowchart of the operation of the envelope extraction circuit 2.

First, enter the frequency Freq of the sampling frequency F _s and tone, calculating the number N of the tone data in the extraction interval (S
1, S2, S3). As a result, data corresponding to the basic period of a musical tone is exactly included in one extraction section.

Further, the length (seconds) of the musical sound is inputted, and the total number M of the extraction sections is obtained from this information, the sampling frequency and the number of data in one extraction section (S4, S5).

The variable j of the extraction section number and the variable i of the data position in one extraction section are initialized to "1" (S6, S7).

The variable MAX of the maximum value and the variable MIN of the minimum value of the musical tone data in one extraction section are initialized to “0” (S8), and by i = N,
Until the scanning within the current extraction section is completed (S13), MAX and
MIN is compared with the current tone data A _ij (the i-th tone data in the j-th extraction section), and the current tone data A
_{If ij} is greater than MAX, update MAX with that value (S9, S1
0), if less than MIN, update MIX with that value (S11, S1
2) Increment i to advance the musical tone data to be compared next (S14). Therefore, when i reaches N, the maximum value and the minimum value of the musical sound data in the extraction section scanned this time are stored in MAX and MIN, respectively.

Then, when i reaches N, the difference between MAX and MIN is entered in the j-th envelope value variable E _j, and j is incremented to advance to the next extraction section (S15, S17). J
Until M reaches M, that is, until the extraction of the envelope value in all the extraction sections is completed.

After completion of the process shown in FIG. 3, the sequence of the variable from E ₁ to E _M represents the M envelope data extracted from tone data.

4 to 7 show waveform examples of musical tone data and waveform examples of envelope data extracted according to the present embodiment. FIG. 4 shows the waveform of the damping sound of the trumpet, and FIG. 5 shows the waveform of the envelope data extracted from the waveform data. FIG. 6 shows the waveform of the continuous sound of the trumpet, and FIG. 7 shows the envelope waveform extracted therefrom.

In the above embodiment, as the extraction section,
A section corresponding to the cycle of the waveform (a basic cycle representing the pitch of the sampled sound) is selected as a section of the waveform data to be inspected to extract one envelope data, and the section of the basic cycle is sampled. Frequency (F
s) / Calculated based on the fundamental frequency (Freq) of the designated and input musical tone, but based on the sampling frequency data and the designated and input fundamental frequency value data of the sampling tone, a section corresponding to another predetermined period (for example, a musical period May be selected as an extraction unit section.

[Effects of the Invention] As described in detail above, in the present invention, in order to extract envelope data from digital waveform data of a sound digitally sampled at a predetermined sampling frequency, a fundamental frequency of the sampled sound is used. The basic frequency value data representing the digital waveform data is determined by a predetermined section length determining means from a value of the sampling frequency and the basic frequency value data input and specified, and a data section of a predetermined cycle related to the digital waveform data is determined. In equal division means, the entire waveform section of the digital waveform data is equally divided into small sections having the length of the data section of the predetermined cycle,
The maximum / minimum value determining means calculates the maximum value and the minimum value of the waveform data included in each small section, and the difference calculating means calculates the difference between the maximum value and the minimum value of each small section, and envelopes The envelope value of each small section is determined according to the difference calculated for each small section by the value determining means. Therefore, the asymmetry of the waveform of the sampled sound can be compensated, and the extraction section (the section of the waveform data to be inspected to extract one envelope data)
Has the advantage of guaranteeing sections of equal length. Therefore, envelope data can be reliably and accurately extracted from a sampled musical sound without losing its musical characteristics.

[Brief description of the drawings]

1 is an overall configuration diagram of an envelope extraction device according to an embodiment of the present invention, FIG. 2 is a configuration diagram of a main part of an envelope extraction circuit, FIG. 3 is a flowchart showing the operation of the envelope extraction circuit, and FIG. FIG. 5 is a waveform diagram showing an example of a tone waveform, FIG. 5 is a waveform diagram showing an envelope waveform extracted according to the embodiment based on the tone waveform of FIG. 4, FIG. 6 is a waveform diagram showing another example of a tone waveform, FIG. 7 is a waveform chart showing an envelope waveform extracted according to the embodiment based on the waveform of FIG. 1 ... tone data memory, 2 ... envelope extraction circuit, 3 ... extraction envelope memory, 21 ... extraction section setting scanning section, 22 ... section musical tone data reading section, 23 ... maximum / minimum extraction section.

Claims (2)

(57) [Scope of request for utility model registration] An envelope extracting apparatus for extracting envelope data from digital waveform data of a sound digitally sampled at a predetermined sampling frequency, comprising: inputting fundamental frequency value data representing a fundamental frequency of a sampled sound; Frequency data input means, predetermined section length determining means for determining a data section of a predetermined cycle related to the digital waveform data from the sampling frequency value and the basic frequency value data, and all waveform sections of the digital waveform data, Equal division means for equally dividing into small sections having the length of the data section of the predetermined period; and for each of the small sections equally divided by the equal division means, the maximum value and the minimum value of the waveform data included in the small section. Means for determining the maximum / minimum value for determining the value; Envelope extraction apparatus characterized by comprising difference calculating means for calculating the difference between the value and the envelope value determining means for determining an envelope value for each small section in accordance with the difference calculated for each small section. 2. The envelope extracting apparatus according to claim 1, wherein said predetermined section length determining means determines a basic period of a sampled sound from the sampling frequency value and the basic frequency value data. An envelope extraction device, wherein a data section having a length corresponding to the predetermined period is determined as the data section having the predetermined period.