JPS58165442A - Encoded storage reproducing device - Google Patents

Abstract

PURPOSE:To obtain easily a reproducing signal with good quality from digital data having the reduced amount of data with a simple constitution, by changing the pass band of a low pass filter in response to the sampling period set at every one frame signal. CONSTITUTION:A digital signal outputted from an AD converter ADC is a 8- bit digital signal, in which the inputted sound signal is sampled and quantized at a prescribed sampling period (1/8000 sec in the example of this explanation) at all times, and stored in the 1st storage device M1 (the 1st memory of a buffer memory, M1) under the control of a control circuit CCT sequentially. In an encoded storage reproducing device, an analog signal obtained by DA-converting the sampling train corresponding to frame signal read out sequentially from the 2nd memory M2 is given to a low pass filter VLPF of a variable pass band which is variably controlled with a control signal obtained by DA-converting a sampling period TC in the sampling train, allowing to prevent the component of strain being a frequency >=1/2 of the sampling frequency from being produced in the reproducing signal.

Description

[Detailed Description of the Invention] When an audio signal is encoded and transmitted or generated as a digital signal, conventional methods for reducing the amount of data as much as possible include logarithmically compressing the signal amplitude or calculating the difference. It is well known that various methods have been adopted, such as modulation or delta modulation, but these conventional methods seek to reduce the amount of data in the amplitude direction, resulting in quantization distortion. Therefore, there was a problem that the quality of the two-way signal deteriorated.

When the amount of data decreases significantly, the applicant's company does not determine the decrease in bits in the amplitude direction, but rather in the time axis direction, and uses the waveform obtained from the sample value sequence and the original. We proposed an approximation compression method for audio signals that can be expected to significantly reduce the amount of data by approximating the waveform so that the degree of difference from the signal waveform is less than a certain ratio (Japanese Patent Laid-Open No. 5
(Refer to Publication No. 6-155998), and achieved some results, but in this existing proposed method, the sampling period is determined based on the deviation of the signal waveform from the previous Himoto basket, and the signal waveform Since it was not determined based on the signal itself, there was a tendency for it to be insufficient in terms of reproduction of fine waveforms in the signal.

In order to solve the problems with the existing proposal of the age table by the applicant company, the applicant company uses a sampling period such that the interval between the ze points of the signal is approximately equally divided. ``1'' We have proposed a modern digital encoding device that encodes signals. By the way, while the digital encoding device described above can easily encode signals, the sampling period is random. Therefore, if a distortion component caused by an error from the original signal is included in the reproduced frequency band when the signal is generated internally, the problem is that it cannot be removed and is included in the reproduced signal. In other words, if a signal is encoded with a sampling period in which the zeρ point interval is divided into approximately equal parts for each zep point interval of the signal, then in the signal part with the length of the zep point interval, Naturally, the sampling period becomes longer, but the signal part with the longer sampling period (I
According to the sampling theorem, the frequency band in which the reproduced signal corresponding to the low frequency part of the I main frequency is distortion-free is a frequency band that is less than 1/2 of the sampling frequency. The passband of the low-pass filter is set to a poorly determined sampling period of 4 x 5, which corresponds to the part with the shortest zero point interval in the signal to be encoded. Since the distortion component in the Okao signal corresponding to the long-period signal portion may be included in the passband of the low-pass filter installed in the reproduction system, The problem is that signal distortion inevitably increases.

The present invention has been made for the purpose of providing an encoding storage/reproduction device capable of solving the above-mentioned problems, and is an object of the present invention to provide an encoding storage/reproduction device that can encode a signal to be encoded at intervals of a predetermined period of time. signal (l-frame signal), and performs encoding such that the sampling period for each l-frame signal is determined according to the state of change in the signal on the time axis in each one-frame signal. Digital data is a set of a sample value indicating the amplitude value of the signal and data indicating the sampling period for obtaining the original value, and the sampling frequency is set for each frame of the signal during playback. Using a variable pass band type low pass filter whose pass band is controlled so that the pass band is less than half the frequency band, the pass band of the variable pass band type low pass filter is expressed as the sampling period in digital data. By controlling based on data, it is possible to easily obtain a reproduced signal free of harmonic distortion.Hereinafter, with reference to the attached drawings, specific details of the encoding storage and reproducing apparatus of the present invention will be described. will be explained in detail.

Figure 1. 1 is a block diagram of an example of the configuration of a Kiyu Yusei device that includes the Hokugo storage and amphibious device of the present invention, and in FIG. MIC is a microphone. ρhon,
LPF is a low frequency Uraha device. ADC is a pulse converter, CG is a clock pulse generator, C<'T is a control circuit including a microcomputer, OP is an operation unit, and Br
is Kikama button, Bp is Amyosei button, and Bs is Stop button. Also,
Peng is the first storage device (first memory) %M, is $I12
security device (second memory), DAC, . DAC is a transducer, VLPF is a variable pass band type low-pass filter, DAC is an amplifier, and SP is a speaker.

FIG. 2 is a waveform example diagram of the signal 8m before encoding,
The line 00-0 in Figure 1K2 is the AC axis line shown for reference. In the waveform diagram shown in Figure 2, the signal f is 8 m
is the time length f of the signal portion divided into fixed time lengths on the time axis, and each signal portion of the above-mentioned time length Tf is . Each signal is called one frame signal.

The digital data to be decoded 9 in the encoding storage/reproduction device of the present invention has a sampling period for each one frame signal depending on the state of change on the time axis of the signal in each one frame signal. The coding as determined is carried out, and the coding mode is such that it has a specific relationship with the number of ZE points in the signal of one frame, and If the sampling period is determined and the frequency I spectrum of the signal of 1 frame is 1'':2, then the level will be reduced by a constant ratio of 1'':2. The frequency of the component with the highest frequency value among the components within the range before decreasing @
A case in which the sampling period for each frame is determined so as to have a specific relationship with respect to WC, and . There is. ．．・So, first of all, like the former... If encoding is performed using a coding mode, the following will be explained.・Second
The illustrated signal 8aK 《1. Predetermined time length Tf
Each frame signal having a signal crosses the AC axis 0-0 line multiple times within the time length Tf, that is, it has a plurality of zero points within the time length Tf. However, the number of ZE points in the signal of each frame ▲ differs depending on the frequency components of the signal in each frame, for example. #2! To explain the waveform Ss shown in , there are eight zep points in the signal of one frame from time t, to t, and
In the signal of one frame from 2 to t, there are four Ze points, and below, for the signal κ of successive l frames on the time axis, the number of Zep points is 6, 3, There are 4 pieces i
It turns out that

Now, as shown in #, the number is related to the number of Zep points present in the predetermined time period Tf signal portion of the signal Sm, that is, the signal of each 1 frame, and the signal of K%l frames. The sampling period T such that the time length is equally divided by
e, a code ~.・ {Evtrfx゜(!:, f7.,, #lo#$1E;
The following is an explanation of this point, taking as an example the signal S shown in #!2 shown above.

In other words, as mentioned above, the number of zero points in the signal of one frame successively on the time axis, such as the signal 8m shown in the diagram $2, is
If the number of zero points is 8, 4, 6, 3, or 4, for example, for an l-frame signal with 8 zero points, the time length Tf is divided into (8XK) equal parts. For example, for a one-frame signal with four zero points, the time length Tf is (4XK. ), K, the same goes for Tlc, and the number of zero points for each l frame is 6 or 3 so that a sample value sequence from the signal of that l frame can be obtained with a sampling period that is divided into equal parts. Regarding the signal, a sequence of sample values from each frame of the signal is obtained at a sampling period such that the time length Tf is equally divided into (6XK) or (3XK), and generally, The number of zero points in the signal of one frame is 2.
In the case of
A sequence of sample values can be obtained, and by using an encoding means such as ``Shoji'', it is possible to easily reduce the recording and reproducing operation while minimizing the amount of data.

The data obtained by an encoding means such as #Ill', that is, the Sakemoto value sequence from each one-frame signal having a predetermined time length Tf, is one frame! The data obtained by sampling with the sampling period obtained by equally dividing the time length Tf by a number (ZXK) that has a specific relationship with the number of zero points 2 in the signal of one frame is , the data is combined with information such as the sampling period Te, the number N of sampled values in a signal of one frame, and the frame number, and is used for transmission or storage.

Next, referring to the block diagram of FIG. 1, the encoding operation described in [III] will be explained.

The microphone p-phone MIC shown in FIG.

- In the sickle internal generation device shown in the first diagram, a microphone σ phone MIC is used as a signal source, but the signal source may be a generator of other forms of audio signal or a generator of other signals. .

In the description of the embodiments below, the cutoff frequency of the low frequency filter LPF is assumed to be 3KHs.

The low-pass filter LPF [Thus, the audio signal converted into a signal in the frequency band of 3 KHz or less is converted into a digital signal with the required number of bits (8 bits in the following explanation) by the conversion converter ADC, and then sent to the microphone p. It is supplied to the control circuit OCT which is configured including a computer, and the inverter converter ADC mentioned above is supplied with 8K from the Kutsk pulse generator CG.
The conversion is performed by pulses with a repetition frequency of Hm.

The digital signal output from the D converter M°C is the input audio signal that is always converted to a constant sampling period of k (1/8000 m in the example) and then quantized to 8
It is a digital signal of bits, which is sequentially stored in the storage device M, @1 (first memory M, or buffer memory Ml) under the control of the control circuit OCT. #i
The puffer memory M1 is assumed to have a storage capacity of 51i bytes in the following explanation example, and it is divided into two parts each having half the storage capacity, and the two parts are alternately filled with data. Used for writing and reading data.

Now, the encoding recording operation of #1l illustrated device is as follows: Operation 11
By operating the recording hammer button Br at 0P, the operation is performed according to the Pugler^ as shown in the flowchart shown in FIG.
When r is operated, Pugura A starts (at the beginning of 1 in the figure 3), and in step 《1》
The 9-bit sample counter, the 8-pit zero point angle, the 16-bit frame counter, and the like provided in the register are reset.

Even before the key pot button Br is operated, that is, at the age of "F Hajime" in the first feature boat shown in the third diagram, the first
The control circuit CCT of the illustrated recorder is
By receiving pulses from the generator CG of G.
``I am performing an interrupt operation of step (to) and
}Puffer memory M1 digital signal output from M℃
are stored sequentially, and a 9-bit sample counter is counted up.

Also, from the beginning onwards, the 9-bit sample counter is reset and repeated each time it reaches a full count.

At step (+), the stored sample value is read from the buffer memory M, and the 9-bit sample counter is counted up by one.

In step <<3>>, it is checked whether the sign of the sample value read in step (2) is the same as the sign of the sample value immediately before it, and if there is no change in sign, it is determined that the sample value is at the zero point. If there is a change in sign, the sample value read in step (2) is assumed to be a zero point, and the process proceeds to step (4). In step (4), the 8-bit zero point is Count up the counter by 1.

In step <<5>>, check whether the number of sample values sequentially read from the buffer memory M1 has reached 256 (or 512) using the count value of the 9-bit sample counter.
/<The number of sample values read from sofa memory M is 25
When you reach 6K (that is, repeat steps (2) to (4) 25 times)
If the process is repeated six times), proceed to step (6), and if the number of single values consecutively output from the buffer memory M has not reached 255, return to step (2).

Here, the number of sample values 256 read out from the buffer memory M1 as described above is determined by the D converter M° being a constant sample value for the signal 8m shown in FIG. The number of sample values obtained by sampling at a period of (.1/8000 seconds) is 256!
This corresponds to the time length Tf of the 2-ray signal.

In step (6), using the count value Zc of the 8-bit Zee point counter, the predetermined number K, and the number 256 representing the time length Tf of the l-frame signal, the sample in the l-frame signal is calculated. Calculate the period ↑ctl- required to obtain the value sequence, and also calculate the number of samples N.

Machine conversion period Tc = 256/Zc@K Number of squares pJ = 256/Tc Next, in step (7), a sample value sequence is extracted from the buffer memory M by applying the regular conversion period Tc. In order to sequentially read sample values for each sampling period Tc regarding the signal of the frequency A, the count value of a 9-bit sample counter (address counter) every Tc is used as an address signal from the buffer memory M. Sequential K
Read the N sample values and create data that combines the frame number of the count value Fc of the 16-bit frame counter, the number of samples N5, the sampling station period Tc, and the N sample values with the age recorded. , it is stored in the storage device Mt (second memory M2) K of the second home, and the process proceeds to step (8).

In step (8), the 16-bit frame counter is counted up by 1.

In step (9), check to see if the 16-bit frame counter has reached a full count or if the stop button Ba has been operated. If it's in the autumn state, then it's over, and if not, go to step (2)? Go back and check each step above. ′ Repeat step.・1: ・・::・1′ Sub-buffer memory M, which has a storage capacity of 512 pies as described in 1, is divided into 2 parts with a storage capacity of 1/2, as described above. 2% is used alternately for writing and reading, and the signal time of l frame is 32 ms, l
Assume that a signal with 256 samples in 7-ray A is being stored and read out, and now, for example, the number K12 mentioned above is
If stipulated in rc. Assuming that the number Ze of Ze points in the signal of one frame A is 32, the sampling period Te is Tc=256/32X2=4, that is, 478000=0.5 (milliseconds).

In the case of the above example, the number N of squares is 64, and the signal of frame A, which consists of 2560 samples and has a time length of 32 milliseconds, has a number N of squares 64. This sample size is 64
If a low frequency filter is used, for example 750 Hz, no distortion will occur in the two-way signal.

Also, the average of the point interval 2 in I for the entire signal is 32
What if it was? , frame counter count number F
2″17..ite corresponding to c, 1 corresponding to the number of samples N
byte, 1 byte corresponding to the sampling period Tc, and a 64-byte Kusumoto value string. 6 for 1 frame signal
Since the memory M of #I2 with a storage capacity of 8 bytes is required, the memory M of M2 is now required. &L If a 64K byte memory is used, 963 frames, or a signal of about 30 seconds or more, will be stored in the second memory M2.

As is clear from the explanation so far, the second memory M2
K is a digital signal that is a set of data with a sampling period Tc, a sequence of sample values, data on the number of camphor trees N, 7-ray 8 number Fc (calculated value Fc of the frame A counter), etc. for each one-frame signal. is stored, but the amount of data is greatly reduced compared to the original digital signal stored in the first memory M, (buffer memory M,). The above-described encoding performed at the Kama time reduces the amount of data and allows recording of audio signals over long periods of time with a small memory capacity.

Next, if the encoding mode is the latter described above, that is,
The sampling period in the signal K of each l-frame is equal to l-frame▲
A case will be described in relation to the frequency spectrum of the signal.

In this case as well, in the device shown in Figure 1, the audio signal output from the microphone MIC is transmitted to the converter ADC [given by Converter A performs conversion using pulses with a repetition frequency of 8 KHz from a pulse generator CC.
The DCWC converts the signal into a digital signal with the required number of bits (8 bits in the following explanation), and supplies the digital signal to the control circuit CCT, which includes a microphone computer, and the digital signal is processed under the control of the control circuit CCR. It is the same as the case described above that the data is stored in the puffer memory M1.

Now, in the following explanation example, the buffer memory M, is assumed to have a storage capacity of 512 heights, and is divided into two storage areas each corresponding to the storage capacity, and the two memory areas are It is assumed that it is used for sequentially writing data for signals of alternating Vct frames and reading data for signals of l frames.

That is, the two storage areas in the buffer memory M1 each have a storage capacity of 256 bytes, and each 8-bit (1 byte) sample value sent from the converter ADC every t7'sooo seconds. l consisting of 256 pieces of
The data to be compared with the frame signal is stored in the buffer memory M
While the data is being stored in the first storage area in the puff memory M1, data corresponding to the signal of frame ^ that was already stored therein is read out from the second storage area in the puff memory M1. Based on the data read from the second storage area, the control circuit OCT performs calculations to obtain the frequency spectrum of the signal of the frame corresponding to the read data, and calculates the peak level of the obtained frequency spectrum. White'. 'j,,'F#i
k; *y) The point at which the pitch level decreases is defined as the threshold level, and the level below that threshold level. Detection of the highest yaw wave number component among the above components tBLrs9”: Detected.
i]:' The time length corresponding to the reciprocal of the frequency value that is approximately twice the frequency value of the frequency value component is set as the sampling synchronization Tc, and data every sampling period Te is set from the storage area of the block 2 in the buffer memory M, The sample values of 7 dresses are read out one after another to create the signal data of frame A consisting of a new Tatemoto value sequence, and then KI
Data related to the sampling period Tc described in IIJ and other necessary data are added to create a set of data, and the data is stored in the second memory M2.

After the above encoding is performed on the i frame of the I frame stored in the second storage area in the buffer memory M, this time, for the storage area of $2 in the buffer memories M and K, Ω The signals of frames t and ot of the converter ADC are newly stored, and the signals of frame l stored in the first storage area of the buffer memories M and rc are read out, and the above-mentioned processing is performed. A similar process is performed to create l-frame data consisting of a new sample value sequence for the l-frame signal, and the sampling period T used to obtain the new column sequence.
A set of data is added with data related to e, other necessary data, etc., and is stored in the memory M of stripe 2.

As described above, the signals of each successive frame on the time axis are sequentially encoded and sequentially stored in the second memory M. Next, Jllg
The recording operation in the apparatus shown in FIG. 1 in this case will be described with reference also to the flowchart shown in FIG. #
I1 The recording sickle operation of the device shown in FIG.
When the l button Br is operated, the operation is performed according to the boograph shown in the fu-chart shown in FIG. 4. When the record sickle button Br on the operation section OPvc is operated,
When the ρ program starts (1 start in FIG. 4), a 9-bit sample counter, a 16-bit 7-frame counter, etc. provided in the control circuit cc'r are reset in step (1r).

1111. Since the record button Br is operated. That is, the "beginning" I1lIvC in the F11 chart shown in Figure 4
In this case, the control circuit CCT of the device shown in the first diagram receives pulses from the Kutsk pulse generator CG, so that
While performing the interrupt operation in step (8r), the digital signal output from the Ω conversion-Nκ is sent to the buffer memory and the K
The 9-bit Hiromoto counter is also counted up.

Also, in the beginning of "E Hajime", the 9-bit Sakemoto counter is repeatedly reset each time it reaches a full count.

Step (2r) read out the Kusumoto value stored in the buffer memory M, and calculate the frequency spectrum of the l-frame signal by Fourier transformation;
Then, the power spectrum of each order signal component is calculated and the process proceeds to step (3r)K.

Calculation of the frequency spectrum and power spectrum of the signal of one frame of Aki may be performed by using all 256 sample values corresponding to the signal of frame A at once, but it is possible to increase the calculation speed. In order to do this, the 256 sample values corresponding to the signal of 1 frame are divided into 8 groups, each group consisting of 32 machine values read out consecutively on the time axis, and each group is divided into 8 groups. Calculate the frequency spectrum and power spectrum for each group. It is better to arithmetic average the obtained calculated values to obtain the frequency spectrum and power spectrum of the l-frame signal, and the explanation of the following example will also be based on 256 sample values corresponding to the l-frame signal. This is done for the case where the calculation is divided into 08 groups of 32 groups.

The harmonics when the frequency spectrum is calculated by dividing the 256 sample values corresponding to the l-frame signal into 8 groups each consisting of 32 sample values read out consecutively on the time axis. The highest degree of is l6.

Next, in step (3r), the calculated power spectrum P-'z'*冫;"e""* and the power spectrum P≦C, *":=":';";;, μ below (
For example, excluding signal components with a level of 1/64 or less, the highest harmonic component among the remaining signal components is detected.
II, II Take out and advance step (4r) tic.=. :step (
4r), the next ll11l! corresponds to the order of the harmonic component outputted from #ae. [Determine the sampling period Tc by referring to the table shown.

The unit of the numerical value shown in the column of sampling period in Table 1 is the sampling period (1/8000 seconds in the description of the embodiment) in the converter ADC, as described above. ), and in the range of the highest order from 6 to l6, the original Rihon conversion period is 1.2 and 11111. ．． 1 Correctly corresponding orders are 'il for sampling period l.
ll"""5?1'.7, the order is 16, and for the sampling period 2, the order is only 8 (
．． However, in the table shown in Table 1, the first orders 9 to 16 are made to correspond to the sampling period l, and the orders 6 to 8 are made to correspond to the sampling period 2.

Note that it is clear from the sampling theorem that the highest frequency of the signal to be generated is 1/2 of the sampling frequency shown in Table 1.

In step (5v), in order to sequentially read the sample values for each sampling period Tc determined as described above from the buffer memory M1, the count value v7 of the 9-bit sample counter (address counter) is Tc. As the address signal, KN sample values are sequentially read from the buffer memory VM, and the count value FcO frame number of the 16-bit frame counter, the number of samples N, the sampling period Te, and the age are written as y: Create data that is paired with N sample values, and use it as #
Stored in memory M of l2, step (6r) I/C
move on.

In step (6r), it is checked whether the 16-bit 7-frame counter has reached a full count or whether the stop button Bs has been operated, and if the frame counter has reached a full count or the stop button B8 has been operated. If so, the process is over, and if not, return to step (2r) and repeat each of the above steps.

2 bytes for the count number FcVc of the frame counter, 1 byte for the number of samples N, 1 byte for the sampling period Tc, and a sample value string of 64 bytes Wc,
A second memory with a storage capacity of 68 bytes is required for every signal of ll frames, but if a 64K byte memory is used as the second memory M, then the second memory M, [ 963 frames, that is, a signal of about 30 seconds or more will be stored.

As is clear from the explanation so far, the second memory M,
For each frame of signal, K stores digital data in which data of sampling period Tc, sample value sequence, data of number of camphor trees N, frame number Fc, etc. are set. Is the amount of data significantly reduced compared to the digital data stored in the buffer memory Mt? - By the above-mentioned encoding, the amount of data is reduced, and a small capacity memory 47 (thus, it becomes possible to store audio signals for a long time). Furthermore, if the signal is encoded based on the frequency spectrum of each l-frame signal as described above, it is clear that the quality of the reproduced signal will be improved compared to the former case described above. .

Next, a case where the signal stored in the second memory M2 as described above is read out and an audio signal is generated by the apparatus of the present invention will be explained with reference to the flowchart shown in FIG. do.

The program shown in FIG. 5 starts ("Start" in FIG. 5) by operating the automatic button Bp on the operating section of the device shown in FIG. , in step (2P), the number of samples N in the signal of l frame and the data of the sampling period Te are read from the second memory M, and then in step (3P)
P), the sample value is stored from the second memory M, I.
I is read out one by one and given to the biconverter DAC, and the data of the sampling period Tc is read out one by one to the biconverter DAC. device DAC.

``?゜'? In step (4P), a time wait is performed so that the time required for one round of steps (2P) to (5P) coincides with the time length indicated by the sampling period Tc.

1/8000 = (Te-1) (Poogram loop time)
1+{step (2P). Step (3P). Step (5P) and fixed part of step (4P)} In step (5P), it is determined whether reading of N sample values has been completed, and if it has not been completed yet, step (2P) ), and if completed, go to step (6P).
).

In step (6P), check whether the frame counter has reached a full count or whether the stop button Bs has been operated, and if the frame counter has reached a full count or the stop button Ba has been operated, it is over, or whether not. If so, return to step (2P).

Bi converter DAC,I in the above step (4P)
DAC,”-Ode?.::Ji,7ge-lfl supply allI
From the two ICs "C" and the DAC, the analog signals corresponding to the successive Kumoto values within the signal of one frame are input to the variable passband type low-pass wave filter VLPF as an input signal. Further, from the Dk converter DAC, an analog signal corresponding to the sampling period Tc of the sample value in the l-frame signal is given to the variable pass band type low voltage filter VLPF as its control signal.

FIG. 6 shows ■Converter DAC1. It is a book circuit diagram showing a configuration example of the DAC part and the low-pass filter VLPF in the variable passband area, and the low-pass filter VLPF in the variable passband area.
F has a higher cut-off frequency in its variable range of cut-off frequencies than resistors R, . R, and capacitor C, . C, and the lower cutoff in the variable range of cutoff frequency. ．． The frequency is determined by the resistors R, , R, and the capacitors c, *cx, and the frequency value in the middle of the cut-off frequency variable range is determined by the DA converter D.
The 7-pin switch, which is switched by the data input of the sampling interval Tc input to the AC, and the variable resistor connected in series with each other also control the bird, and input the data of the sampling interval Tc. The frequency values are respectively adjusted to predetermined degrees.

The sampling period Te and . An example of the correspondence between the sampling frequency fs corresponding to the sampling period Te and the cutoff frequency fc of the area filter me is shown in the following jIz table.

Variable passband type low-pass filter VL shown in FIG.
PF is the sampling period given to the DA converter DAC↑
Depending on the data value of C, the gain of amplifier A changes depending on the resistance selected by the 7-nag switch ASW, and it becomes κ.
Therefore, the photocoupler PC, . PC, [photodiode Pd. The cutoff frequency changes as a result of changes in the resistance values of the photosensitive resistors (for example, Cd8) (2) and (2) caused by changes in the amount of light emitted by Pd, and the pass band is variably controlled. In addition, in the iris 6 diagram, RleR4mB@~R. is a resistance, -, the bird is a variable resistance! i-At, A2 is an amplifier, C
、． C, is a capacitor, pClmPC! is a photocoupler.

As is clear from the above explanation, in the encoding storage/reproducing apparatus of the present invention, the signal of each frame ^ sequentially read out from the memory M of the library 2 and the corresponding sample value sequence are obtained by translating the sequence of sample values. The analog pg signal is applied to a variable passband low-pass filter VTPR whose passband is variably controlled by a control signal obtained by performing Te1klm conversion for the first period of sampling in the sampled value sequence, and is input into the reproduced signal. This can be done so that distortion components with a frequency value higher than 1/2 of the sampling frequency are not generated.

In this way, in the encoded storage/reproduction pupil of the present invention,
The signal to be encoded is divided into signals (l
digital data obtained by encoding U such that the conversion period for each frame is determined according to the state of change on the one-time axis of the signal in each frame signal 11. tli
l; 9 layers per signal of 7 frames 1. 1.? ．． The passband of the low-pass wave filter is changed in accordance with the sampling period set for each l-frame signal so that the harmonic components in the angular raw signal are outside the passband. Therefore, it is clear that all of the above-mentioned problems can be satisfactorily solved by the device of the present invention, and a high-quality reproduction signal can be easily obtained from digital data with a reduced amount of data using a simple configuration. can.

[Brief explanation of drawings]

The St diagram is a Pook diagram of an embodiment of the sickle angle generation device including the encoded storage internal generation device of the present invention, JIII2
The figure is an explanatory waveform example diagram, IJN3 to IJN5 are explanatory flow charts, and Figure 6 is a variable passband type low-pass filter and a Poutsk 1-way diagram of its related parts. . MIC...Microphone, LJ'F...Low frequency ura wave device, . ADC...converter, CG...clock pulse generator, C'CT--q, control circuit including computer, OP--1? :Operation unit, Br・
・・Record button, 11 Bp-・・＾Raw button, 1, :',,. Stop button, M1...
l-th memory device (buffer memory iJ)-M*...second
storage device, 'DAC, . DAC,・Ilm converter, V
LPF/Mf passband type low frequency wave generator, -220=

Claims (1)

[Claims] The signal to be encoded is a signal for each predetermined time length (17 frame signal), and each frame is Digital data obtained by encoding such that the sampling period for the frame signal is determined, that is, a sample value indicating the amplitude value of the original signal, and data indicating a sampling period tube for obtaining the sample value. A code storing and reproducing device for decoding digital data, such as a set of sample values indicating the amplitude value of the recorded original signal and sampling for obtaining the sample values. The sample value indicating the amplitude value of the original signal in the digital data that is paired with the data indicating the period is corrected on the time axis, then converted, and converted into a variable passband type low-frequency converter. Means for supplying the signal as an input signal to Haichi Ura, the Kuzumoto value indicating the amplitude value of the original signal distributed in May, :, s, and the data indicating the sampling period for obtaining the sample value are made into a set. For data indicating the sampling period in digital data that is
Means for applying the A conversion to the variable pass band type low pass filter as a control signal is provided, and the cutoff frequency of the variable pass band type low pass filter is changed by the k control signal. A coding storage endogenous device which can obtain an i endogenous signal without harmonic distortion by