JPH039478B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a musical sound waveform processing device for an electronic musical instrument, which is used to digitally process musical sound waveforms having periodicity.

[Prior art]

2. Description of the Related Art Techniques for sequentially sampling analog musical tone signals and converting and processing them into digital sampling data have been widely used in the field of electronic musical instruments in recent years.

By the way, a particularly important point in this type of technology is how to reduce the number of bits of digital data to be handled (how to compress the data), and various data compression methods have been considered for this purpose. .

FIG. 7 is a diagram showing the general configuration of a data compression circuit. In this figure, 1 is an input terminal to which digital sampling data SD is sequentially supplied;
2 is a predicted value generation circuit. This predicted value generation circuit 2 predicts the value of the current sampling data SD from the past sampling data SD, and outputs this prediction result as a predicted value Y to the subtracter 3. The subtracter 3 subtracts the predicted value Y from the current sampling data SD, and outputs the subtraction result as data D to the coder 4. Here, the predicted value Y is close to the current sampling data SD, and therefore,
The value of the output data D of the subtracter 3 is much smaller than the value of the sampling data SD, and the number of bits thereof is also smaller. That is, the predicted value generation circuit 2
Data compression is performed by a data compression circuit 5 comprising a subtracter 3 and a subtracter 3. The compressed data D is then converted into, for example, a Huffman code, a Shannon Juano code, etc. in a coder 5, thereby further compressing the data and outputting it from an output terminal 6. The data output from the output terminal 6 is transmitted to another location or stored in a memory, for example.

FIG. 8 is a diagram showing the configuration of a demodulation circuit that demodulates data compressed by the circuit of FIG. Therefore, it is returned to data D, and then returned to sampling data SD by a demodulation circuit 10 comprising an adder 9 and a predicted value generation circuit 2.

FIG. 9 is a diagram showing the simplest configuration example of the data compression circuit 5 shown in FIG. 7, and the data compression method using this circuit is DPCM (Differential Pulse Code
Modulation). In this circuit,
Reference numeral 12 is a delay element for delaying input data by one sample time, and for example, a D-type flip-flop is used. FIG. 10 shows the configuration of a demodulation circuit that demodulates data compressed by the circuit of FIG. 9. Hereinafter, the circuits in FIGS. 9 and 10 will be referred to as a DPCM compression circuit 13 and a DPCM demodulation circuit 14, respectively.
It is called. By the way, the original signal (musical tone signal, etc.)
When is a signal with a smooth waveform, greater data compression efficiency can be obtained by connecting the circuit shown in FIG. 9 in multiple stages as shown in FIG. Note that FIG. 12 shows the demodulation circuit of FIG. 11. However, if the number of stages in the circuit shown in FIG. 11 is larger than a certain level, greater compression efficiency can be obtained by configuring the data compression circuit based on the so-called linear prediction method. FIG. 13 shows its configuration, and FIG. 14 shows the configuration of the demodulation circuit. In addition, in these figures, M is a multiplier, a1
~an is the multiplication factor (value to be multiplied). The multiplication coefficients a1 to an are obtained by solving a normalization equation based on an autocorrelation sequence of input data.

[Problem that the invention seeks to solve]

Each data compression circuit described above can effectively compress data, but in the field of electronic musical instruments, etc., the amount of data handled is extremely large, so it is desired to further improve compression efficiency. There is.

SUMMARY OF THE INVENTION It is therefore an object of the present invention to provide a musical sound waveform processing device for an electronic musical instrument that can compress the amount of data even more than the above-mentioned conventional device when processing musical sound waveforms having periodicity.

[Means to solve the problem]

First invention: delay means for outputting delayed signals obtained by delaying a musical sound waveform having periodicity by 1 to n cycles; multiplication means for multiplying each of the delayed signals by a coefficient; and arithmetic means for subtracting each of the multiplied delayed signals, and outputs the output of the arithmetic means as a processed signal.

Second invention: A musical sound waveform processing device for an electronic musical instrument that returns the processed signal obtained by the first invention to the original signal, and includes an adding means and an output of the adding means that is delayed by 1 to n cycles. and a multiplication means for multiplying each output of the delay means by a coefficient, and the addition means adds the processed signal and each output of the multiplication means and outputs the result.

Third invention: A plurality of signal compression circuits are connected in cascade, each comprising a delay means for delaying a signal supplied to the input end by a predetermined period, and a subtraction means for subtracting the output of the delay means from the signal supplied to the input end. It is composed of

Fourth invention: A musical sound waveform processing device for an electronic musical instrument that returns the compressed signal to the original signal according to the third invention,
The demodulation circuit comprises an adding means and a delay means for delaying the output of the adding means by a predetermined period and supplying the delayed output to the input end of the adding means, and the same number of demodulation circuits as the signal compression circuits are connected in cascade.

〔Example〕

FIG. 1 is a block diagram showing the configuration of a data compression circuit to which a signal processing device 16 according to an embodiment of the first invention is applied. In this figure, T1 is an input terminal, and digital sampling data SD obtained by sampling musical tone signals at regular time intervals is sequentially supplied to this input terminal T1. In this case, the number of sample points during one period of the musical tone signal is m. The sampling data SD supplied to the input terminal T1 is compressed by the DPCM compression circuit 13 and supplied to the signal processing device 16 as data D1. FIG. 2 is a diagram showing an example of sampling data SD and data D1 using analog waveforms.

The signal processing device 16 is a circuit that compresses the data D1, and is composed of a predicted value generation circuit 17 and a subtracter 22. In the predicted value generation circuit 17, 18
-1 to 18-P are delay elements that respectively delay input data by m sample times (that is, one period of the musical tone signal) and output the delayed data, and for example, m-stage shift registers are used. Data D1 is set to 1 by these delay elements 18-1 to 18-P.
~P periods are delayed and supplied to multipliers 19-1 to 19-P. Each of the multipliers 19-1 to 19-P is
The input data is multiplied by coefficients b1 to bP and output to adders 20-1, 20-2, . . . Adder 2
0-1, 20-2... are multipliers 19-1 to 19-
Add each output of P, and use the addition result as the predicted value Y1
It is output to the subtracter 22 as . The subtracter 22 subtracts the predicted value Y1 from the data D1 and outputs the subtraction result to the coder 4 as data D2. coder 4
(see FIG. 7) converts the code of data D2 and supplies it to output terminal T2 as data D3. Data D3 obtained at this output terminal T2 is stored in a memory or transmitted to another location.

In the above configuration, the data D1 has periodicity as is clear from FIG. 2, and changes at a period equal to the period of the musical tone signal. Therefore,
The current data D1 is replaced with each data D from 1 to P cycles ago.
It is possible to predict from 1. The predicted value generation circuit 17 is a circuit constructed from the above considerations,
By appropriately determining the coefficients b1 to bP of the multipliers 19-1 to 19-P, it is possible to generate the predicted value Y1 very close to the current data D1. In addition,
The larger the number of delay elements 18, the closer the predicted value Y1 to the current data D1 can be generated. Then, the closer the predicted value Y1 is to the current data D1, the more highly compressible data can be obtained as the data D2.

Next, a method for determining the coefficients b1 to bP will be explained. The coefficients b1 to bP are determined by statistical methods. Now, if each value of the data D1 is y ₀ , y ₁ , y ₃ ... y _n , y _n+1 , y _n+2 ..., the signal processing device 16 calculates y ₀ , y _n , y _2n . ...〓y ₀ ,l y ₁ ,y _n+1 ,y _2n+1 ...〓y ₁ ,l 〓 〓 y _n-1 ,y _2n-1 ,y _3n-1 ...〓y _n-1 ,l It acts independently on each of the m time series. Therefore, let the autocorrelation sequences of y ₀ , l ~ y _n-1 , l be r ₀ , l ~ r _n-1 , l, respectively, and let r _ln-1 〓 ^k=0 r _k , l ...(1) By finding a new autocorrelation sequence and solving the normal equation based on this, the coefficients b1 to bP can be found.

That is, first, an autocorrelation sequence is obtained as follows. First, decompose the input sequence y _o as follows.

y _0,l(l=0,1, … _L) y ₀ ,y _n ,y _2n …y _Ln y _1,l(l=0,1, … _L) =y ₁ ,y _n+1 ,y _{2n +1} …y _Ln+1 〓 〓 y _n-1,l(l=0,1, … _L) =y _n-1 ,y _2n-1 ,y _3n-1 …y _{(L+1)n- 1} (2
) Next, calculate the autocorrelation sequence for each, and let the sum be the true autocorrelation r〓 _o .

r _i,o = _Lo 〓 ⁱ⁼⁰ y _i,l・y _i,l+o (i=0,1,...m-1)...(3) r〓 _o = _n-1 〓 ⁱ⁼⁰ r _i,o ...(4) The coefficients b1 to bP that minimize the data D2 can be obtained by solving the following normal equation.

Note that the above procedure is equivalent to the following procedure. First, find r _o based on the formula r _o = _N 〓 ^m=0 y _n , y _n+o (6). Next, r _o is determined based on the following formula: r _o = r _no (n = 1, 2...P) (7).

Furthermore, since it is known that the linear sum of positive definite autocorrelation sequences is positive definite, the demodulation system using the coefficients b1 to bP (see FIG. 3) is also stable.

Next, an embodiment of the second invention will be described. FIG. 3 is a diagram showing the configuration of a demodulation circuit that returns the data D3 compressed by the circuit of FIG. 1 to the original sampling data SD. In this figure, T
3 is an input terminal, and data D3 is sequentially supplied to this input terminal T3 in the same order and at the same timing as when it was output from the output terminal T2 in FIG. This data D3 is then inversely converted into data D2 by the decoder 8, and successively supplied to the signal processing device 30 according to the embodiment of the second invention. The signal processing device 30 converts the data D2 into data D
1, and includes an adder 31 and a predicted value generation circuit 17 having exactly the same configuration as that in FIG.
It is composed of. The basic configuration of this signal processing device 30 is the same as the demodulation circuit consisting of the adder 9 and predicted value generation circuit 2 shown in FIG. , are supplied to the DPCM demodulation circuit 14. The DPCM demodulation circuit 14 demodulates the data D1 into sampling data SD and supplies it to the output terminal T4.

Next, an embodiment of the third invention will be described with reference to FIG. In FIG. 4, sampling data SD supplied to input terminal T1 is compressed by DPCM compression circuits 13, 13, and supplied as data D4 to signal processing device 40 according to the third embodiment of the invention. This signal processing device 40 further compresses the data D4 and outputs it as data D5.
A plurality of compression circuits 43 each consisting of a subtracter 41 and a delay element 42 having m sample times are connected in cascade. Here, the original signal (musical tone signal, etc.)
If the data D4 has periodicity, the data D4 also has periodicity. Therefore, the delay element 42 operates as a predicted value generation circuit, and the compression circuit 43 can compress the data. And this compression circuit 4
Data compression can be further achieved by cascading a plurality of 3. The data D5 output from the signal processing device 40 is code-converted by the coder 4 and is supplied to the output terminal T2 as data D6.

Next, an embodiment of the fourth invention will be described with reference to FIG. FIG. 5 is a diagram showing the configuration of a demodulation circuit that demodulates data D6 compressed by the circuit of FIG. , are supplied to the signal processing device 45 according to the embodiment of the fourth invention. The signal processing device 45 includes a demodulation circuit 4 comprising an adder 46 and a delay element 47 with m sample times.
8 are cascaded in a number equal to the number of compression circuits 43 in FIG.
Demodulate and output. The output data D4 is
The data is returned to the original sampling data SD by the DPCM demodulation circuits 14, 14, and is supplied to the output terminal T4.

In the circuits shown in FIGS. 1 and 4 described above, the digital sampling data SD is compressed, but the DPCM compression circuit 13 and signal processing devices 16 and 40 are configured with analog circuits, and the original analog The signal may be data-compressed by these circuits and then converted into digital data.

Next, an application example of the above-described embodiment will be described. FIG. 6 is a block diagram showing an example of the configuration of an electronic musical instrument to which the circuit shown in FIG. 3 is applied. In this figure, the memory M stores data (hereinafter referred to as musical tone data) obtained by compressing sampling data obtained by sampling the entire waveform of a musical tone signal from its rise to its end using the circuit shown in FIG. It is stored for each tone (piano sound, flute sound, etc.) and for each pitch. When one of the timbres is selected by the timbre selection section 50, the output data TC of the timbre selection section 50 specifies the storage area of the memory M in which musical tone data corresponding to that timbre is stored. be done. Next, when any key on the keyboard 51 is pressed, the key press detection section 52 detects this and outputs the key code KC of the pressed key and a key-on signal KON. This key-on signal KON remains “1” while the same key is pressed.
Continue signal. The address generator 53 converts the key code KC output from the key press detection unit 52 into corresponding address data AD1 and outputs it to the memory M, and also from the time when the key-on signal KON rises to the “1” signal. , 0, 1, 2, . . . address data ADD that changes sequentially is output to the memory M. When address data AD1 is supplied to memory M,
An area corresponding to the address data AD1 within the storage area specified by the data TC mentioned above is specified. This area is an area in which musical tone data having the timbre selected by the timbre selection section 50 and having the pitch of the key code KC output from the key press detection section 52 is stored. Then, when the address data ADD is supplied to the memory M, the tone data is sequentially read out starting from the relative address 0 of the area, and is sequentially supplied to the tone reproduction section 54.
The musical tone reproduction section 54 is the same circuit as in FIG. 3,
When musical tone data is sequentially supplied, sampling data SD is outputted from the output terminal and supplied to the sound system 55. The sound system 55 is
The supplied sampling data SD is converted into an analog signal and produced as a musical tone from a speaker.

Note that if pitch changes such as vibrato are to be added to the generated musical tone, the output timing of the address data ADD may be changed. In this case, data processing within the musical tone reproduction section 54 is also synchronized with the change timing of the address data ADD. In addition, instead of storing musical tone data for each pitch, if you want to store one set of musical tone data for all pitches, or one set of musical tones and data for each pitch, use the address Change timing of data ADD,
The speed may be set to correspond to the key code KC.

〔Effect of the invention〕

As explained above, according to the first to fourth inventions,
When a periodic signal is digitally processed, the amount of digital data (number of bits) to be processed can be significantly reduced compared to the conventional method.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of a data compression circuit to which an embodiment of the first invention is applied, and FIG.
FIG. 3 is a block diagram showing the configuration of a data demodulation circuit to which an embodiment of the second invention is applied, and FIG. 4 is a diagram showing an example of the data SD and D1 in the figure using analog waveforms. Block diagram showing the configuration of a data compression circuit to which one embodiment is applied, No. 5
The figure is a block diagram showing the configuration of a data demodulation circuit to which an embodiment of the fourth invention is applied, FIG. 6 is a block diagram showing an example of application of the circuits of FIGS. 1 and 3, and FIG.
The figure is a block diagram showing the general configuration of a data compression circuit, FIG. 8 is a block diagram showing the configuration of a data demodulation circuit corresponding to FIG. 7, and FIGS. 9 and 10 are respectively
11 is a block diagram showing a configuration example of another data compression circuit. FIG. 12 is a block diagram showing the configuration of a demodulation circuit corresponding to FIG. 11. FIG. 13 is a block diagram showing the configuration of a data compression circuit based on the linear prediction method, and FIG. 14 is a block diagram showing the configuration of a demodulation circuit corresponding to FIG. 13. 16, 30, 40, 45...signal processing device, 1
7... Predicted value generation circuit, 18-1 to 18-P...
Delay element, 19-1 to 19-P...multiplier, 20
-1, 20-2...Adder, 22...Subtractor, 3
1...Adder, 41...Subtractor, 42...Delay element, 43...Data compression circuit, 46...Adder,
47...Delay element, 48...Data demodulation circuit.

Claims (1)

[Scope of Claims] 1. A musical sound waveform processing device for an electronic musical instrument that processes a musical sound waveform having periodicity, comprising: delay means for respectively outputting delayed signals obtained by delaying the musical sound waveform by 1 to n cycles; and each of the delayed signals. and a calculation means for subtracting each delayed signal multiplied by the coefficient from the musical sound waveform, and outputting the output of the calculation means as a processed signal. Musical sound waveform processing device for electronic musical instruments. 2 Delay a periodic musical sound waveform by 1 to n cycles to obtain a plurality of delayed signals, multiply each of these delayed signals by a coefficient, and then subtract the resulting processed signal from the musical sound waveform. A musical sound waveform processing device for an electronic musical instrument that returns the musical sound waveform to a musical sound waveform, comprising: an adding means; a delay means for outputting a signal obtained by delaying the output of the adding means by 1 to n cycles; and multiplying each output of the delay means by a coefficient. A musical sound waveform processing device for an electronic musical instrument, further comprising a multiplication means, and the addition means adds the processed signal and each output of the multiplication means. 3. A musical sound waveform processing device for an electronic musical instrument that processes a musical sound waveform having periodicity, including a delay means for delaying a signal supplied to an input end by a predetermined period, and an output of the delay means from a signal supplied to the input end. A musical sound waveform processing device for an electronic musical instrument comprising a plurality of signal compression circuits connected in cascade, each comprising a subtracting means for performing subtraction. 4 A plurality of signal compression circuits connected in cascade, each comprising a delay means for delaying a signal supplied to the input end by a predetermined period, and a subtraction means for subtracting the output of the delay means from the signal supplied to the input end. A musical waveform processing device for an electronic musical instrument that returns a signal compressed by the device to the original signal, comprising an adding means, and a delay for delaying the output of the adding means by a predetermined period and supplying the delayed output to the input end of the adding means. A musical sound waveform processing device for an electronic musical instrument, comprising the same number of demodulation circuits as the signal compression circuits connected in cascade.