JP2621466B2 - Sampling device - Google Patents

Description

Description: TECHNICAL FIELD [0001] The present invention relates to a sampling device having a loop reading function of repeatedly reading waveform data representing a sampled acoustic signal.

[Prior art and its problems] Conventionally, an electronic musical instrument having a sampling function has been known. For this type of sampling electronic instrument,
Since the memory capacity is limited, in order to generate a continuous sound, a loop reading function for repeatedly reading waveform data stored from a specified loop start address to a loop end address of the memory is provided. I have.

In a sampling electronic musical instrument having a loop reading function of this type, conventionally, two different waveform data are repeatedly and simultaneously read out from a memory, and arithmetic processing is performed so that each of the data is faded in and faded out from each other. In some cases, a so-called cross-fade loop process is performed to obtain

However, according to such cross-fade loop processing, since two waveform data having different phases and amplitudes are added, interference occurs with each other, and as a result of cancellation, a change occurs in the volume level, which is accompanied by the loop processing. There is a problem that a phenomenon called "sound fluctuation" occurs repeatedly at a predetermined loop cycle.

In order to effectively perform such cross-fade loop processing, a section suitable for loop processing, that is, a section where good sound quality can be expected, must be set as a loop section from the waveform data sequence. Since waveform data over a long period of time is required, there is a problem that storage means having a large memory capacity is required.

Also, like various percussion sound signals,
In the case of waveform data where the level of the attack part is large and sharp, and the level change of the release part is sharp, the loop section is generated by setting the loop section to an attack part containing many harmonics of many orders in order to improve the S / N ratio. The musical tone has a hard sound quality containing many harmonic components. Conversely, in the case of waveform data related to an acoustic signal of an attenuation system such as a piano,
Since the loop section is set to the sustain part or the release part in order to obtain the characteristics of the sound quality, the low-level waveform data is loop-processed, and the generated tone has a poor S / N ratio related to quantization noise. Problem.

[Object of the Invention] The present invention has been made in order to solve the above-described problems, and does not generate unnecessary sound fluctuation when generating a musical sound by loop-reading out waveform data representing a sampled sound. It is another object of the present invention to provide a sampling device capable of adjusting sound quality and improving an SN ratio due to quantization noise.

[Summary of the Invention] In order to achieve the above object, the present invention reads out stored waveform data from a loop start address to a loop end address of a storage means by a reading means, and obtains a spectrum of the read waveform data. Is extracted by the first calculating means, the extracted spectrum is stored in the spectrum storing means, the stored spectrum is corrected by the correcting means, and the corrected waveform data is corrected by the second calculating means based on the corrected spectrum. Generating means for reading out from the predetermined address in the storage means to an address immediately before the loop start address by the reading means, and combining the read waveform data with the corrected waveform data generated by the second arithmetic means. Add to each other at a predetermined ratio that changes with time Synthesizing to generate synthesized waveform data, the reading means reading the waveform data from the start address of the storage means to the address immediately before the predetermined address, the synthesized data generated by the synthesis means, Second
The main point is that the musical tone corresponding to the corrected waveform data repeatedly generated by the calculating means is instructed to be generated by the musical sound generation instructing means.

Further, the present invention further comprises a corrected waveform data storage means for storing the corrected waveform data generated by the second arithmetic means, wherein the reading means sets a predetermined address of the storage means to an address immediately before the loop start address. And the modified waveform data read from the modified waveform data storage means are added and combined with each other at a predetermined ratio that changes with time by the synthesizing means so as to instruct generation of synthesized waveform data. The point is.

Further, the present invention is characterized in that the correction means has a display means for displaying the spectrum extracted by the first calculation means.

Further, the present invention further comprises an address setting means,
The point is that at least one of the loop start address and the loop end address or the predetermined address can be set arbitrarily.

Embodiment An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows the overall circuit configuration of a sampling electronic musical instrument according to an embodiment of the present invention. In the drawing, a CPU 1 is a central processing unit of a microcomputer and controls various operations of the electronic musical instrument according to a program stored in a ROM 2. I do. The tone signal input from the microphone 3 is amplified by the amplifier 4, the frequency band component equal to or higher than half the sampling frequency is cut off by the filter 5, and is added to the A / D converter 6. Under the sampling start instruction from the CPU 1, the A / D converter 6 samples the tone signal input through the filter 5 at a certain level or higher for a predetermined time, and the digitally converted waveform data is stored in the memory M1 in the RAM 7. .

When the operator of the operator unit 8 is operated, under the control of the CPU 1,
The entire waveform data stored in the memory M1 is displayed as a waveform on the display screen 9-1 of the display unit 9 composed of, for example, a liquid crystal panel (see FIG. 2). The display screen 9-1 displays the waveform
Dotted lines L ₁ and L ₂ are displayed. These dotted lines L ₁ and L ₂ correspond to the addresses A ₁ and A ₂ of the memory M1, respectively, and are used for discrete extraction for spectrum extraction as described later.
The section for performing the Fourier transform (hereinafter referred to as DFT) is displayed.

Doing DFT performed by the operation of the operator section 8, CPU1 waveform data stored in from the memory M1 from the address A ₁ to the address A ₁ -1 is read by, for example, a rectangular window by computing function of CPU1 window A function-based DFT is performed. The data of the amplitude spectrum and the phase spectrum obtained as a result of the DFT (the data of the phase spectrum is initial value data) is stored in a predetermined area of the memory M2, and the amplitude spectrum is stored in the display screen 9-1 of the display unit 9 on the display screen 9-1. It is displayed as shown in FIG. In FIG. 3, the second
Those denoted by the same reference numerals as those in the figure have the same function, and the spectrum at the left end shows the DC component of the waveform data, and the spectrum on the right side is the fundamental tone when the DFT section is regarded as one cycle ( The spectrum of the first harmonic is shown, and the spectrum of the harmonic corresponding to each order is shown below.

FIG. 4 is a view showing a type 4 of the operation unit in the operation unit 8.
An example of the arrangement of the cursor keys C1 to C4 on a panel image (not shown) of the musical instrument is shown. By moving the cursor C-2 on the display screen 9-1 shown in FIG. 2 left and right by turning on the cursor key CK1 or the cursor key CK2, an arbitrary DFT section and the synthesis of the original waveform and the corrected waveform are performed. You can set a starting point (described below)
Data indicating this interval, the address A _1, is stored in a predetermined area of the memory M2 in RAM7 as the address A _2.
By turning on the cursor keys CK1 and CK2, the cursor C-3 on the display screen 9-1 shown in FIG. 3 is moved right and left to select an arbitrary spectrum whose amplitude value is to be corrected. Then cursor CK3 and cursor CK4
Is turned on to increase or decrease the amplitude value while viewing the display screen 9-1.

Subsequently, the reverse discrete /
When the execution of Fourier transform (hereinafter referred to as IDFT) is instructed, the IDFT is executed by the arithmetic function of CPU1, and the waveform data of the same section as that of the previously executed DFT is stored in a predetermined area of memory M2 as corrected waveform data. Is done.

Here, in order to increase the SN ratio of the waveform data stored in the memory M2 with respect to the quantization noise, the number of bits of the memory M2 may be used to the maximum limit, and the maximum value m of the amplitude of the waveform data is determined by the number of sampling bits. And the maximum value V of the signal level at which sampling can be input determined from the above. For this purpose, the maximum value m of the amplitude of the waveform data after the execution of the IDFT is temporarily stored, the amplitude value of each amplitude spectrum is multiplied by V / m, and the IDFT is executed again.

Appears as shown in Figure 2 when the display example on the display unit 9 the waveform data after execution of the IDFT, in the displayed waveform, the beginning portion of the waveform to the dotted line L ₁ is stored in the memory M1 and it is those corresponding to the original waveform data and, in the section from the dotted line L ₁ to the dotted line L ₂ corresponds to the corrected waveform data described above. At the time of reproduction, the position of the cursor C-2 determines at what timing the waveform data after IDFT stored in the memory M2 is read (whether to sound) as described above.
The position of -2 can be arbitrarily set by operating the cursor keys CK1 and CK2 of the operating section 8. Also this position is intended to correspond to the address A ₀ of the memory M1, the setting value is stored in a predetermined area of the memory M3 in RAM 7. Further, it is assumed that various parameters for generating a musical tone, such as envelope data for performing envelope processing on waveform data read from the memories M1 and M2 by the CPU 1, are stored in the memory M3 in advance.

The keyboard 10 is used by a player to perform a performance operation by designating a desired pitch.When the performance operation is performed, the CPU 1 stores the waveform data stored in the memories M1 and M2 in accordance with the pitch specification. Read at speed.

CPU1 first before one of the addresses A ₀ described above from the start address of waveform data stored in the memory M1, that is sequentially read up to the address (A ₀ -1), the read waveform data is, control of the CPU1 (For example, from the start address to the address (A ₀ -1),
After multiplying the envelope value by the MAX value, the signal is applied to the D / A converter 11.

The CPU1 is subsequently before one of the address A ₀ described above in the waveform data stored in the memory M1 of the address A _1,
That address (A ₁ -1) to the address from the start address of the waveform data stored in the memory M2 (A ₂ -
A ₁ ) and are read simultaneously. Here, the waveform data read from the memory M1 is latched by the latch circuit L1, and the waveform data read from the memory M2 is latched by the latch circuit L2. The waveform data from the latch circuits L1 and L2 are input to multipliers 12 and 13, respectively, multiplied by the count data from the CPU 1, and then added by the adder. The multiplication and addition processing will be described later, but is for crossing one waveform to the other and adding envelope waveform data. Further, subsequently CPU1 reads repeated (from the head address address A ₂ -A ₁₎ waveform data stored in the memory M2 to.

After the tone signal analog-converted by the D / A converter 11 is subjected to a predetermined filtering process by the filter 15,
The sound is emitted as a musical sound through the amplifier 16 and the speaker 17.

Next, the operation of the embodiment having the above configuration will be described focusing on the operation of the arithmetic processing on the waveform data read from the memories M1 and M2 under the control of the CPU 1.

2 read from memory M1 and memory M2 respectively
Multipliers 12, 13 and adder 14
The arithmetic processing is performed based on the following equation.

W (A) = E (t) W ₁ (A) · (1−X (A)) + E (t) W ₂ (A−A ₁ ) · X (A) where W (A) is the address A , W ₁ (A) is the value of the waveform data at address A of the original waveform data stored in the memory M1, and W ₂ (A−A ₁ ) is the IDFT stored in the memory M2. address when the start address of waveform data after execution was 0 (a-a ₁₎ the value of the waveform data in, X (a) is, X (a) = (a -a 0) / (a 1 -A ₀ ) where A ₀ <A <A ₁ . E (t) is the envelope data at time t. Here, W ₂ (A−
In A ₁ ), A−A ₁ is a negative value, but the value of W _{2 at} a negative address is W ₂ (−1) = W ₂ (A ₂ −A ₁ −1) W ₂ (−2) = W ₂ (A ₂ −A _1-2 ) is defined as a loop.

Here, X (A) is represented by a straight line as described above. However, in general, any function can be used as long as X (A ₀ ) = 0 and X (A ₁ ) = 1. A value calculated by an equation such as
Cross-fading different waveform data read from M2, that is, gradually fade out the original waveform data read from one memory M1, and gradually fade in the corrected waveform data read from the other memory M2 Thus, a process for naturally connecting two different waveform data is performed.

Usually, when you want smoothly connect the corrected waveform data as the original waveform data, it sets the position of A ₀ to the attack portion,
If you do not want to damage the attack part of the original waveform data,
Sustain part of the position of A _0, it may be set to release part.

As described above, in the above embodiment, the sampled original waveform data and the predetermined section of the original waveform data are
After correcting the spectrum and cross-fading the corrected waveform data obtained by IDFT,
The modified waveform data obtained by IDFT is read out in a loop. In other words, different waveform data is not read out by cross-fading in a loop, but only the corrected waveform data is looped, so that the loop operation does not cause interference between the two waveform data. In addition, unnecessary sound fluctuation does not occur. In addition, by modifying the original waveform data, for example, adjusting the sound quality by cutting harmonic components, and further improving the S / N ratio if necessary, that is, the waveform data suitable for the loop operation can be obtained, so that the loop can be obtained. There is no need to store a large amount of waveform data due to a long sampling time in order to obtain waveform data suitable for the memory device. Therefore, the memory capacity of the memory M1 for storing waveform data representing a sampled sound can be reduced.

In the above embodiment, after the DFT is performed on the original waveform data stored in the memory M1, only the amplitude spectrum is corrected, but the phase spectrum can be corrected. In general, when the phase spectrum is changed, the phase of each harmonic changes, and when the original waveform data and the corrected waveform data are mixed, the amplitude of each harmonic increases or decreases due to interference and changes unnecessarily. In an extreme case, if the phase of a certain harmonic is reversed between the original waveform and the corrected waveform, the amplitude of the harmonic may cancel each other and become zero. Further, even if the phase is not inverted, the amplitude of each overtone is unnecessarily increased or decreased due to the phase shift, thereby causing sound fluctuation.

However, by appropriately correcting the phase spectrum, it is effective, for example, when looping the waveform data of a non-integer harmonic. That is, for example, if 10 cycles of the basic cycle are set as the waveform data to be looped, the loop waveform data includes 1/10 harmonics, 2/10 harmonics in addition to the integer harmonics.
Includes (N + n / 10) harmonics. If the original combined data contains, for example, 2.91 harmonics, it is considered that the sound quality hardly changes even if the original combined data is replaced with 2.9 = (2 + 9/10) harmonics. When the DFT processing is performed once on the waveform data in the loop section, the 2.91 harmonic is decomposed into the 2.9 harmonic and other harmonics, but the other harmonics can be removed in principle by modifying the amplitude spectrum.

In this way, it is possible to replace the overtone of 2.91 harmonics in the attack part and 2.9 harmonics in the loop section,
Nevertheless, since the sound quality is changed in the portion that enters the loop section, in the above-described embodiment, such a change in the sound quality is prevented by performing mixing by crossfading. However, the difference between the 2.91 harmonic and the 2.9 harmonic (difference in frequency) may cause a “buzz” at the cross portion of the two waveform data. As a modified embodiment of the present invention, the phase spectrum is also corrected by correcting the phase spectrum. It is conceivable to minimize the "beat" for each overtone.

In the above embodiment, the spectrum extracted by performing DFT on the original waveform data is corrected, and the waveform data resulting from the execution of IDFT is stored in the memory M2. If possible, the IDFT processing can be performed in real time when a tone signal is generated, that is, during a performance. In this case, it is sufficient to store only the spectrum data after DFT execution in the memory M2,
Since it is not necessary to store the waveform data after execution of IDFT in the memory M2, the memory capacity of the memory M2 can be reduced.

Further, in the above embodiment, the DFT is performed on the original waveform data, and the IDFT is further performed on the result. However, the present invention is not limited to such a Fourier transform or inverse Fourier transform, and a spectrum is obtained from the original waveform data. May be executed, as long as it is an arithmetic method that can generate the original waveform data from the spectral components.

In the above embodiment, the present invention is applied to a keyboard instrument. However, the present invention is not limited to a keyboard instrument, and may be an electronic string instrument,
The present invention is applicable to all types of sampling electronic musical instruments such as electronic wind instruments.

[Effect of the Invention] As described above, according to the first aspect of the present invention, the readout means reads out the waveform data stored from the loop start address to the loop end address of the storage means, and obtains the spectrum of the read waveform data. Is extracted by the first calculating means. The extracted spectrum is stored in the spectrum storage means, and the stored spectrum is corrected by the correction means. The corrected waveform data is generated by the second calculating means based on the corrected spectrum. The reading means reads out from the predetermined address in the storage means to the address immediately before the loop start address, and combines the read waveform data with the corrected waveform data generated by the second arithmetic means by the synthesizing means. The combined waveform data is generated by adding and synthesizing each other at a predetermined ratio that changes. Based on the waveform data read from the start address of the storage means to the address immediately before the predetermined address by the read means, and further based on the synthesized data generated by the synthesis means, The musical tone is generated by repeating the corrected waveform data generated by the calculating means.

Therefore, when a musical tone is generated by loop-reading waveform data representing a sampled sound to generate a tone, waveform data obtained by mixing different waveform data is not loop-read, so that interference of different waveform data causes a loop cycle. Unnecessary sound shaking due to repeated selling does not occur. In addition, since the corrected waveform data obtained by correcting the spectrum of the base waveform data is read out in a loop to generate a musical tone, according to the correction of the amplitude spectrum, the sound quality can be adjusted by cutting harmonic components, or the level can be adjusted. In addition to improving the S / N ratio by raising the level of low waveform data, a sampling device that can reduce the "beat" for each harmonic by further correcting the phase spectrum is obtained.

Further, the invention according to claim 2 further comprises a corrected waveform data storage means for storing the corrected waveform data generated by the second arithmetic means, and one of the predetermined address or the loop start address of the storage means by the reading means. The waveform data read up to the previous address and the corrected waveform data read from the corrected waveform data storage unit are added and synthesized by the synthesizing unit at a predetermined ratio that changes with time to generate synthesized waveform data. I did it.

According to a third aspect of the present invention, the correction means has a display means for displaying the spectrum extracted by the first arithmetic means, so that the player can check the display on the display means with his / her eyes. , The spectrum can be easily corrected.

Further, the invention according to claim 4 further comprises address setting means, so that at least one of the loop start address and the loop end address or the predetermined address can be arbitrarily set. Since it is possible to arbitrarily set a section in which the stored waveform data and the corrected waveform data are added and synthesized, and further, it is possible to arbitrarily set a section in which the corrected waveform data is looped.
A desired tone can be obtained more easily.

[Brief description of the drawings]

The drawings are for explaining a sampling electronic musical instrument according to one embodiment of the present invention, and FIG. 1 is a diagram showing the overall circuit configuration thereof, and FIG. 2 is a display screen 9- of the display unit 9 in FIG.
1 is a diagram showing waveform data displayed on FIG. 1, FIG. 3 is a diagram showing an amplitude spectrum also displayed on the display screen 9-1, and FIG. 4 is a cursor key on the operation unit 8 in FIG.
It is a figure showing CK1-C4. 1 ... CPU, 3 ... Microphone, 6 ... A / D converter,
8 ... operation part, 9 ... display part, 9-1 ... display screen,
11 D / A converter, 12, 13 Multiplier, 14 Adder, M
1, M2, M3 …… Memory.

Claims (4)

(57) [Claims] 1. A storage means for storing waveform data representing a sampling sound, wherein said waveform data stored from a loop start address to a loop end address specified by said storage means is repeatedly read by a reading means, In a sampling device having a loop reading function for generating a corresponding musical tone, the reading means reads the waveform data stored from the loop start address to the loop end address of the storage means, and reads the read waveform data. First to extract the spectrum possessed by
Calculating means, a spectrum storing means for storing the spectrum extracted by the first calculating means, a correcting means for correcting the spectrum stored in the spectrum storing means, and a spectrum corrected by the corrected spectrum. A second calculating means for generating corrected waveform data by calculation; and a read-out means for reading from a predetermined address in the storage means to an address immediately before the loop start address. A combination means for adding and synthesizing the corrected waveform data generated by the second arithmetic means at a predetermined ratio which changes with time to generate synthetic waveform data; Generates a tone based on the waveform data read out to the previous address. And a tone generation instruction means for generating a tone based on the synthesized waveform data generated by the synthesis means, and instructing a tone generation by repeating the corrected waveform data generated by the second arithmetic means. A sampling device characterized by the above-mentioned. 2. The apparatus according to claim 1, further comprising a modified waveform data storage means for storing the modified waveform data generated by said second arithmetic means, wherein said synthesizing means uses said read means to read a loop start address from a predetermined address of said storage means. Of 1
The combined waveform data generated by adding and synthesizing the waveform data read up to the previous address and the corrected waveform data read out from the corrected waveform data storage means at a predetermined ratio that changes with time are generated. The sampling device according to claim 1, wherein: 3. The sampling apparatus according to claim 1, wherein said correction means has display means for displaying a spectrum extracted by said first calculation means. 4. The apparatus according to claim 1, wherein said sampling device further comprises address setting means for arbitrarily setting at least one of said loop start address and said loop end address or said predetermined address. Sampling device.