JPH0974358A - Method and device for compressing/expanding digital signal - Google Patents

Abstract

PROBLEM TO BE SOLVED: To satisfactorily compress/expand a ΣΔ-modulated digital audio signal. SOLUTION: The ΣΔ-modulated digital audio signal outputted from a comparator 4 is supplied to a register 7(15) and a symbol (n/j, k) is supplied to a ROM 8. Besides, the signal in the register 7 is supplied to a register 9, and a preceding symbol (m/i) is supplied to the ROM 8. A conversion table (dictionary) prepared by respectively allocating codes (x) of short bits to symbols having high generation probability is written in this ROM 8. Then, the code (x) from this ROM 8 is supplied through a register 10 to a selector 11. Further, when these symbols (m, n/i, j, k) are combined as prescribed, at this selector 11, the code from the register 10 is selected but in the other case, the prescribed code (x) is first selected and the signal from the register 7 is selected later. Thus, data are compressed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital signal compression / expansion method and device used, for example, in the case of digitizing an audio signal by ΣΔ modulation and transmitting (recording / reproducing) the signal.

[0002]

2. Description of the Related Art As a method of digitizing and transmitting (recording / reproducing) an audio signal, a recording / reproducing apparatus such as a CD or DAT, or a digital audio broadcasting such as satellite broadcasting has been conventionally practiced. In such a digital audio transmission device, conventionally, when digitized, a format such as a sampling frequency of 48 kHz, 44.1 kHz or the like and a quantization bit number of 16 bits or the like was specified.

However, in such a conventional digital audio transmission method, the number of quantization bits of the digital audio signal generally defines the dynamic range of the audio signal to be demodulated. Therefore, for example, in order to transmit a signal with higher sound quality, it is necessary to increase the number of quantization bits from the current 16 bits to 20 or 24 bits. However, once the format is specified, it is not possible to easily increase the number of quantization bits, so that it is not possible to extract an audio signal of higher sound quality from these devices.

By the way, a method called ΣΔ (sigma delta) modulation has been proposed as a method of digitizing a voice signal (Acoustic Society of Japan, Vol. 46, No. 3, (1990), pages 251 to 257, “AD / DA converter”). And digital filter (Yoshio Yamazaki) "etc.).

That is, FIG. 24 shows, for example, 1-bit ΣΔ
A configuration using modulation is shown as an example. In FIG. 24, the input audio signal from the input terminal 91 is the adder 92.
Through the integrator 93. The signal from the integrator 93 is supplied to the comparator 94, is compared with, for example, the midpoint potential of the input audio signal, and is quantized by, for example, 1 bit for each sample period. The frequency of the sampling period (sampling frequency) is 48 kHz, 44.1 kHz of the conventional
, 64 times or 128 times that frequency is used. The quantization may be 2 bits or 4 bits.

This digital signal is supplied to the delay unit 95 and delayed by one sample period. This delayed signal is supplied to the adder 92 through, for example, a 1-bit DA converter 96 and added to the input audio signal from the input terminal 91.
Then, the digital signal output from the comparator 94 is taken out to the output terminal 97. Therefore, according to this ΣΔ modulation,
As shown in the above document, by sufficiently increasing the frequency of the sampling period (sampling frequency),
For example, it is possible to obtain a digital audio signal having a wide dynamic range even with a small number of bits of 1 bit.

[0007]

By the way, when recording or transmitting digitized data in the digital audio signal as described above, the amount of data is compressed due to the restriction of the recording medium or the transmission path. May need to be done. The compression method used in that case must be a lossless compression that can completely restore the compressed data to the original data.

On the other hand, it is known that the amount of digitized data increases in the above-mentioned ΣΔ modulation. That is, for example, the sampling frequency 44.1k of the data format used in the conventional digital audio.
In the above ΣΔ modulation, for example, when the sampling frequency is increased 64 times with respect to Hz and the bit length is 16 bits, the amount of data becomes four times even if the data word length is 1 bit.

Therefore, when recording or transmitting such a digital audio signal, it is essential to compress the data amount. Therefore, for compression of such digitized data, various compression methods such as the Huffman code and the LZ method have been conventionally proposed. However, with such a conventional method, a sufficient compression rate cannot be obtained for the above-mentioned ΣΔ-modulated digital audio signal, or the compression processing is complicated, and thus a sufficient effect cannot be obtained. It was

This application has been made in view of such a point, and the problem to be solved is that in the conventional method and apparatus, a means for satisfactorily compressing and expanding a ΣΔ modulated digital audio signal. Was not obtained.

[0011]

Therefore, in the present invention, a digital signal digitized with an arbitrary small number of bits is divided into a plurality of predetermined bits, and the probability of occurrence of the divided digital signal is obtained. A conversion table necessary for performing compression processing based on the probability is created, and a digital signal is compressed by referring to this conversion table. According to this, a ΣΔ modulated digital audio The signal can be compressed and expanded well.

[0012]

BEST MODE FOR CARRYING OUT THE INVENTION That is, according to the present invention, in a digital signal digitized with a predetermined number of bits, the generation probability of a digital signal divided into arbitrary plural bits is calculated, and the compression processing is performed based on this generation probability. In this case, a conversion table necessary for performing is created, and the digital signal is compressed by referring to the conversion table.

By the way, for example, 1-bit ΣΔ signal data is converted into, for example, b (n), b (n + 1), b (n + 2), b (n + 3), b (n +
4) is expressed by dividing this signal into 4 bits from an arbitrary position, for example, B0 = b (n) * 2 ³ + b (n + 1) * 2 ² + b (n + 2) * 2 + b ( Let n + 3) B1 = b (n + 4) * 2 ³ + b (n + 5) * 2 ² + b (n + 6) * 2 + b (n + 7).

Here, for example, 4 bits of the ΣΔ signal are (0,
If it is 1, 1, 0), it can be expressed as B0 = 0 * 2 ³ + 1 * 2 ² + 1 * 2 + 0 = 6. This is a calculation of a so-called hexadecimal value. In this way, for example,
The bit ΣΔ signal can be converted into 16 values (symbols) from “0” to “f” in hexadecimal notation (hexadecimal notation).

That is, 0 = (0,0,0,0) 8 = (1,0,0,0) 1 = (0,0,0,1) 9 = (1,0,0,1) 2 = (0,0,1,0) a = (1,0,1,0) 3 = (0,0,1,1) b = (1,0,1,1) 4 = (0,1,0) , 0) c = (1,1,0,0) 5 = (0,1,0,1) d = (1,1,0,1) 6 = (0,1,1,0) e = ( It can be represented by: 1,1,1,0) 7 = (0,1,1,1) f = (1,1,1,1)

Therefore, an example of the data of the ΣΔ signal when the actual music signal is input is shown in hexadecimal form in the table of FIG. As is clear from this table, for example, the data of the ΣΔ signal when a music signal is input is “0” or “f”.
There is almost no data such as "5" or "a",
There are many data such as "6" and "c".

Further, when the occurrence probabilities of these data are examined, it becomes as shown in the graph of FIG. 2, and it can be seen that the distribution is considerably biased. This is because the ΣΔ signal represents the original music signal by the density of 1-bit “0” and “1”, so the data of “0” and “f” correspond to the negative or positive maximum level signal, and It is considered that this is because it is difficult to appear in a music signal.

On the other hand, if the ΣΔ signal represents a music signal, it is considered that there is a correlation between the respective symbols of the ΣΔ signal. Therefore, by examining the occurrence probabilities of two consecutive symbols (symbol strings), it is found that the distribution is as shown in the table of FIG.

By the way, it can be considered that this table shows the probability of transition from a certain symbol to a certain symbol.
Therefore, in this example, for example, after the symbol "1", 5
It can be considered that “a” appears with a probability of 0% and “c” appears with a probability of 31%. Furthermore, “b” appears with a probability of 16%, “9” appears with a probability of 3%, and other symbols do not occur in this example.

That is, in this example, the symbols appearing after the symbol "1" are "a", "c", and "b".
There are only four types of "9". Therefore, to express these, 2-bit data is sufficient. By doing so, the above-mentioned 4-bit data can be compressed to half the 2-bit data.

However, in reality, the probability that symbols other than these four types will appear after the symbol "1" is not zero at all. Also in the above example, for example, after the symbol "2", "6""7""9""a""c""b"
Four or more types of symbols of "d" are generated, and in these cases, the above-mentioned 2 bits cannot be compressed.

Therefore, in the present invention, for example, for three symbols having a high occurrence probability, 2-bit compressed data "1" (0,1), "2" (1,0), "3".
The code of (1, 1) is assigned, and for other symbols, a code in which the symbol is connected by 4 bits is assigned after 2-bit data “0” (0, 0).

According to this method, therefore, three symbols having a high probability of occurrence have 4 bits of 2 bits, so that the amount of data is compressed by 50%, but in other cases, 2 bits of (0 , 0) plus 4 bits of the original symbol, resulting in a total of 6 bits, which is 50% redundant. However, if the actual data is concentrated on the three symbols with a high probability of occurrence, the data amount can be compressed as a whole.

Further, in the present invention, the amount of data of the ΣΔ-modulated digital audio signal can be compressed by using a device as shown in FIG. 4, for example.

That is, in FIG. 4, the input audio signal from the input terminal 1 is supplied to the integrator 3 through the adder 2. The signal from the integrator 3 is supplied to the comparator 4, is compared with the midpoint potential of the input audio signal, and is quantized by 1 bit for each sampling period. The frequency during the sampling period (sampling frequency) is 48kHz,
Frequencies 64 times and 128 times that of 4.1 kHz are used.

This digital signal is supplied to the delay unit 5 and delayed by one sample period. This delayed signal is supplied to the adder 2 through the 1-bit DA converter 6 and added to the input audio signal from the input terminal 1. As a result, the comparator 4 outputs a digital signal obtained by ΣΔ-modulating the above-mentioned input audio signal.

Then, the digital signal output from the comparator 4 is supplied to, for example, a 4-bit register 7.
Then, the 4-bit symbol (n) written in the register 7 is supplied to the ROM 8. This ROM8 here
Is, for example, in the above-mentioned three symbols with high probability of occurrence,
A conversion table (dictionary) created by allocating 2-bit codes of (0, 1) (1, 0) (1, 1) is written.

Further, the digital signal output from the register 7 is supplied to the 4-bit register 9, for example. Then, the 4-bit symbol written in the register 9 is supplied to the ROM 8. Therefore, from this register 9,
The 4-bit symbol (m) at the immediately preceding timing is RO
Supplied to M8.

As a result, the ROM 8 outputs a 2-bit code assigned by the combination of the symbol (m) and the symbol (n). That is, for example, according to the above-mentioned table of FIG. 3, when the symbol (m) is “0” and the symbol (n) is “a” (0, 1),
The codes of (1, 0) when "b" and (1, 1) when "c" are output, respectively.

Further, for example, when the symbol (m) is "1" and the symbol (n) is "a" (0, 1), when it is "c" (1, 0), when it is "b" (1, The code of 1) is output respectively. Furthermore, for example, the symbol (m) is "2"
When the symbol (n) is "c" (0, 1), it is "a"
The code of (1, 0) is output when, and the code of (1, 1) is output when “b”.

For example, when the symbol (m) is "3" and the symbol (n) is "3" (0, 1), when it is "5" (1, 0), when it is "4" (1, The code of 1) is output respectively. Furthermore, for example, the symbol (m) is "4"
And when the symbol (n) is "d" (0, 1), it is "c"
When (1, 0), and when "e", the code (1, 1) is output.

For example, when the symbol (m) is "5" and the symbol (n) is "5" (0,1), when it is "9" (1,0), when it is "6" (1, The code of 1) is output respectively. Furthermore, for example, the symbol (m) is "6"
And when the symbol (n) is "6" (0, 1), it is "5"
The code of (1, 0) is output when, and the code of (1, 1) is output when “9”.

Further, for example, when the symbol (m) is "7" and the symbol (n) is "2" (0, 1), when it is "1" (1, 0), when it is "3" (1, The code of 1) is output respectively. Furthermore, for example, the symbol (m) is "8"
When the symbol (n) is "d" (0, 1), "e"
The code of (1, 0) is output when, and the code of (1, 1) is output when “c”.

For example, when the symbol (m) is "9" and the symbol (n) is "9" (0, 1), when it is "a" (1, 0), when it is "6" (1, The code of 1) is output respectively. Further, for example, the symbol (m) is "a".
When the symbol (n) is "a" (0, 1), it is "6"
The code of (1, 0) is output when, and the code of (1, 1) is output when “9”.

Further, for example, when the symbol (m) is "b" and the symbol (n) is "2" (0,1), when it is "3" (1,0), when it is "1" (1, The code of 1) is output respectively. Further, for example, the symbol (m) is "c"
When the symbol (n) is "c" (0, 1), it is "a"
The code of (1, 0) is output when, and the code of (1, 1) is output when “b”.

For example, when the symbol (m) is "d" and the symbol (n) is "3" (0, 1), when it is "5" (1, 0), when it is "4" (1, The code of 1) is output respectively. Further, for example, the symbol (m) is "e".
Then, when the symbol (n) is "5" (0, 1), "3"
When (1, 0), and when "4", the code of (1, 1) is output.

For example, when the symbol (m) is "f" and the symbol (n) is "3" (0,1), when it is "4" (1,0), when it is "5" (1,1). The code of 1) is output respectively. In this way, each of the three symbols having a high probability of occurrence is converted into, for example, a 2-bit code. Then, for example, these 2-bit codes (x) are
For example, it is supplied to the selector 11 through the 2-bit register 10.

In the ROM 8, a predetermined discrimination signal is supplied to the control circuit 12 when the symbol (m) and the symbol (n) are not in the above combination. Further, the signal from the register 7 is the selector 11
Is supplied to. Further, for example, a 2-bit code (0, 0) having data “0” is supplied to the selector 11.

In the selector 11, when the symbol (m) and the symbol (n) are in the above combination, the code from the register 10 is selected, and the symbol (m) and the symbol (n) are other than the above combination. In the case of, the 2-bit code (0, 0) is first selected, and then the signal from the register 7 is selected. In this way, the data is compressed.

Further, the digital signal output from the selector 11 is supplied to the synchronization signal and error correction code (ECC) addition circuit 13, and the synchronization signal and the error correction code are added to the digital signal for each predetermined number of samples. . Then, the digital signal to which the synchronization signal and the error correction code are added is taken out to the output terminal 14 for transmission (recording / reproduction).

FIG. 5 shows an example of the format of the signal output to the output terminal 14. In this example, the digital signal is represented by 4 bits (b ₀ , b ₁ ,
b ₂ , b ₃ ) is taken as one data symbol D ₀ , and is re-divided as data symbols D ₀ D ₁ D ₂ . These data symbols are, for example, 12 data symbols D _{0 to} D ₁₁ as one block as shown in B of the same figure, and for example, synchronization signals S _{0 to} S ₃ and error correction codes P _{0 to} P are provided for each block. ₃ is added.

As a result, this apparatus can detect and correct a transmission error that occurs during transmission (recording / reproduction). Further, in the recording / reproducing apparatus, it is conceivable to subject the data to processing such as interleaving so as to sufficiently deal with burst errors occurring on the recording medium.

Further, FIG. 6 shows the configuration of an example of an apparatus for decompressing the data compressed as described above. In FIG. 6, the signal is supplied from the input terminal 30 to the sync separation and error correction circuit 31, and the above-mentioned error correction code P ₀ is supplied.
The error correction is performed by P ₃ and the transmitted (recorded and reproduced) digital signal is taken out.

The digital signal thus taken out is, for example, through the 2-bit register 32, the ROM 3 in which the inverse conversion table (dictionary) corresponding to the ROM 8 is written.
3 is supplied. Further, the ROM 33 is supplied with a signal from a register 34 in which, for example, a 4-bit symbol (m) at a timing immediately before, which is taken out from a selector 36 described later, is written.

Thereby, from the ROM 33, for example, a 4-bit symbol (m) from the register 32 and a 4-bit symbol (m) from the register 34 at the immediately preceding timing, for example, are inversely converted. The bit symbol (n) is retrieved. The extracted symbol (n) is, for example, a 4-bit register 3
5 to the selector 36.

Further, the code (x) from the above-mentioned register 32 is supplied to the control circuit 37 and the code (x) of (0,0) is discriminated. Then, when this determination is made, the above selector 36 connects the code of (0, 0) from the register 32, for example, 4
Control is performed so that the bit symbol is selected, and at other times, the symbol (n) from the register 35 is selected.

In this way, in the above-described apparatus of FIG. 4, for example, 4-bit data is converted into a 2-bit code (x) whose value is "1" to "3" and a code whose value is "0". In the apparatus of FIG. 6, the symbol (n) is inversely transformed using the code (x) and the decompressed previous symbol (m) in the apparatus of FIG. Is expanded.

Furthermore, the digital signal output from the selector 36 can be returned to an analog signal by passing through a simple analog filter 38, and this analog audio signal can be taken out from the monitor terminal 39. Further, this digital signal is subjected to decimation (decimation) filter 40 by digital processing, for example,
It can be converted into an arbitrary signal format such as CD or DAT.

Therefore, the signal converted into the arbitrary format is the digital recorder 4 of the arbitrary format.
1, a CD, a DAT device 42, a mini disk (MD), a digital compact cassette (DCC) device 43, etc. for recording, or a normal DA converter 44 through a reproduction system of these devices for conversion. An analog audio signal can be taken out from the output terminal 45.

In this way, the input audio signal is ΣΔ modulated and transmitted (recorded / reproduced), and the transmitted (recorded / reproduced) signal is reproduced in an arbitrary signal format. The signal thus ΣΔ-modulated is the same as a conventional CD or DA.
Compared to signal formats such as T, for example, 2 in the audible range
Quantization noise is reduced below 0 kHz, and a wide dynamic range can be obtained.

Therefore, in this method and apparatus, the digital signal digitized with an arbitrary number of bits is divided into predetermined plural bits, the occurrence probability of the divided digital signal is obtained, and the compression processing is performed based on the occurrence probability. It is possible to create a conversion table necessary for performing the above, and refer to this conversion table to compress the digital signal.

As a result, in the past, a sufficient compression ratio could not be obtained for a ΣΔ modulated digital audio signal,
According to the present invention, a ΣΔ-modulated digital audio signal can be satisfactorily compressed and decompressed, although a sufficient effect cannot be obtained due to complicated compression processing.

Further, in the above embodiment, 1 symbol = 4
Although the bits are converted into a 2-bit code and the compression / expansion is performed, it is also possible to convert, for example, 2 symbols = 8 bits into a 4-bit code and perform the compression / expansion.

In this case, for example, as shown by a broken line in the device of FIG. 4 described above, a 4-bit register 15 is provided in the subsequent stage of the register 7, and the symbols (k) (written in the registers 7 and 15 (k) ( The 4-bit code (x) is taken out from the ROM 8 from j) and the symbol (i) written in the register 9. Register 10 is 4
Bit.

Also, for example, in the above-mentioned apparatus of FIG. 6 as well, as shown by the broken line, the register 32 has 4 bits, and the code (x) from this register 32 and the previous symbol ( i) and ROM33
Then, the 8-bit symbols (k) and (j) are extracted. The register 35 has 8 bits.

Then, in this case, in the ROMs 8 and 33, "1" is set for the upper 15 data having a high occurrence probability among the 64 types of data formed of 8 bits.
Codes from “(0,0,0,1) to“ f ”(1,1,1,1) are assigned, and other data is assigned to“ 0 ”(0,0,1).
The data is represented by a total of 12 bits by connecting the original 8 bits after (0, 0).

Further, FIGS. 7 to 22 show examples of the dictionary thus formed. That is, each drawing shows the case where the value is "0" to "f" for the immediately preceding symbol (i). Therefore, in these figures, for example, when the value of the symbol (i) is “4”, the converted values of the 8-bit symbols (j) and (k) are (j) = “d” and (k) =, respectively. In the case of "8", the code (x) is converted to "a" in FIG.

Therefore, for example, in the example of the music signal shown in FIG. 1, the first two symbols (jk) = “6”.
For “5”, since the preceding symbol (i) does not exist, the code (x) = “0” and the original data “6” is added.
5 "are connected and converted. Next 2 symbols (jk) =
For “59”, the previous symbol (i) = “5”
Then, the code (x) = “4” is converted from the dictionary shown in FIG.

The next two symbols (jk) = “39”
Is converted by connecting the previous symbol (i) = “9”, the code (x) = “0” in the dictionary of FIG. 16 and then connecting the original data “39”. Also for the next 2 symbols (jk) = “35”, the previous symbol (i) = “9”, the code (x) = “0” in the dictionary of FIG. It is converted by connecting "35".

The next two symbols (jk) = “55”
Is converted into the code (x) = “1” from the above-mentioned dictionary of FIG. 11 with the immediately preceding symbol (i) = “5”. For the next 2 symbols (jk) = “4d”, the previous symbol (i) = “5”, which is shown in FIG.
Is converted into the code (x) = “8” from the dictionary. Further, for the next two symbols (jk) = “58”, the preceding symbol (i) = “d” is converted into the code (x) = “f” from the dictionary of FIG. 20 described above.

By sequentially performing the conversion in the same manner, for example, in the example of the music signal shown in FIG. 1 described above, the conversion is performed as shown in the table shown in FIG. It can be compressed to 60%. Further, in this example, since all the processing is performed in units of 4-bit symbols, the above-mentioned synchronization signal and error correction code addition circuit 13 and synchronization separation and error correction circuit 31 are provided.
It is also suitable for treatment in.

According to the above-described method and apparatus for compressing / decompressing a digital signal, the probability of occurrence of a digital signal divided into arbitrary plural bits in a digital signal digitized with a predetermined number of bits is calculated, and the probability of occurrence is calculated. By creating a conversion table necessary for performing compression processing based on the above and compressing the digital signal by referring to the conversion table, for example, a ΣΔ modulated digital audio signal can be compressed and expanded satisfactorily. is there.

Therefore, by using the method and apparatus of the present invention, a digital audio signal composed of a small number of bits such as 1 bit, such as ΣΔ modulation, can be used without deteriorating its characteristics. It is possible to transmit (record and reproduce) various signals.

In the above description and examples, the conversion table is created based on the example of the data of the ΣΔ signal when the music signal shown in FIG. 1 is input. May give different data. Therefore, for example, it is conceivable to create a conversion table according to the type of the original signal of the digital signal and switch the conversion table according to the type.

However, in the above-mentioned embodiment, for example, when 2 symbols = 8 bits are converted into a 4-bit code and compression / expansion is performed, the high-order 1 with a high probability of occurrence is generated.
Even if the order of the five data is different, the compression result is not affected. Further, even if the occurrence probabilities of the data other than these 15 data are slightly different, the probability of the data not originally in the dictionary occurring is low, so that there is little influence on the compression of these data as a whole.

Therefore, from the above, it is not necessary to create the dictionaries (conversion tables) required for compression according to the types of the original signals, and if one dictionary is created based on the data of any signal. , Can be applied to all signals.

Further, in the above-mentioned dictionaries shown in FIGS. 7 to 22, most data forming the dictionaries are "0", as is clear from the figures. Therefore, if only data other than "0" is selected as a dictionary and data not included in this dictionary is set to "0", the size of the dictionary itself can be made extremely small.

[0068]

According to the present invention, a digital signal digitized by an arbitrary small number of bits is divided into a plurality of predetermined bits, an occurrence probability of the divided digital signal is obtained, and compression is performed based on the occurrence probability. It has become possible to create a conversion table necessary for processing and to perform compression of a digital signal by referring to this conversion table.

According to this, conventionally, a sufficient compression ratio cannot be obtained for a ΣΔ modulated digital audio signal,
According to the present invention, a ΣΔ-modulated digital audio signal can be satisfactorily compressed and decompressed, although a sufficient effect cannot be obtained due to a complicated compression process.

Therefore, by using the method and apparatus of the present invention as described above, a digital audio signal composed of a small number of bits such as 1 bit such as ΣΔ modulation is used, and its characteristics are not deteriorated. It is possible to transmit (record and reproduce) various signals.

[Brief description of drawings]

FIG. 1 is a diagram for explaining an example of a digital signal compression / expansion method according to the present invention.

FIG. 2 is a diagram for explaining an example of a digital signal compression / expansion method according to the present invention.

FIG. 3 is a diagram for explaining an example of a digital signal compression / expansion method according to the present invention.

FIG. 4 is a block diagram of an example of a digital signal compression / expansion device according to the present invention.

FIG. 5 is a diagram showing an example of a data format for the explanation.

FIG. 6 is a block diagram of an example of a digital signal compression / expansion device according to the present invention.

FIG. 7 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 8 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 9 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 10 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 11 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 12 is a diagram showing an example of a dictionary for the explanation.

FIG. 13 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 14 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 15 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 16 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 17 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 18 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 19 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 20 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 21 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 22 is a diagram showing an example of a dictionary for explaining the situation.

FIG. 23 is a diagram for explaining an example of a digital signal compression / expansion method according to the present invention.

FIG. 24 is a diagram for explaining ΣΔ modulation to which the present invention is applied.

[Explanation of symbols]

 1 Input Terminal 2 Adder 3 Integrator 4 Comparator 5 Delay Device 6 1-bit DA Converter 7, 9, 10 Register 8 ROM with Conversion Table (Dictionary) Written 11 Selector 12 Control Circuit 13 Sync Signal and Error Correction Code (ECC) additional circuit 14 Output terminal for transmission (recording / playback)

Claims (10)

[Claims] 1. A compression / expansion method of a digital signal digitized with a predetermined number of bits, wherein the occurrence probability of the digital signal divided into arbitrary plural bits is obtained, and compression processing is performed based on this occurrence probability. A method for compressing / decompressing a digital signal, in which a conversion table required for is created, and the digital signal is compressed by referring to the conversion table. 2. The method for compressing / decompressing a digital signal according to claim 1, wherein the conversion table is created according to the continuity of the divided digital signal of a plurality of bits, and the divided digital signal of a plurality of bits and immediately before that. And / or a digital signal compression / expansion method for compressing the digital signal by referring to the conversion table from the digital signal immediately after. 3. The method of compressing / decompressing a digital signal according to claim 1, wherein the conversion table is created according to the continuity of the divided digital signal of a plurality of bits, and the divided plurality having a high probability of the continuity. A short bit length code is assigned to the bit digital signal, and a specific code of the above short bit length code and the original digital signal are connected to other divided multi-bit digital signals. A digital signal compression / expansion method in which a code is assigned to compress the digital signal. 4. The method of compressing and expanding a digital signal according to claim 1, wherein the conversion table is used to expand the compressed digital signal. 5. The method for compressing / decompressing a digital signal according to claim 3, wherein the presence or absence of the specific code is determined, and when there is no determination, the conversion table is used to decompress the compressed digital signal. A method for compressing / expanding a digital signal, wherein the digital signal connected to the specific code is output when the determination is present, and these are selected to expand the compressed digital signal. 6. A compression / expansion device for a digital signal digitized with a predetermined small number of bits, which is created based on an occurrence probability of the digital signal obtained by dividing the digital signal into arbitrary plural bits. A digital signal compression / expansion device comprising a conversion table and means for compressing the digital signal divided into arbitrary plural bits by referring to the conversion table. 7. The digital signal compression / expansion device according to claim 6, wherein the conversion table is created in accordance with the continuity of the divided multi-bit digital signal, and the divided multi-bit digital signal and immediately before it. And / or a digital signal compression / decompression device comprising means for compressing the digital signal by referring to the conversion table from the digital signal immediately after. 8. The digital signal compression / expansion device according to claim 6, wherein the conversion table is created according to the continuity of the divided digital signal of a plurality of bits, and the divided plurality of the high continuity probability is obtained. A short bit length code is assigned to the bit digital signal, and a specific code of the above short bit length code and the original digital signal are connected to other divided multi-bit digital signals. A digital signal compression / expansion device that is formed by allocating codes. 9. The digital signal compressing / expanding device according to claim 6, further comprising means for expanding the compressed digital signal using the conversion table. 10. The digital signal compression / expansion device according to claim 8, wherein the means for determining the presence / absence of the specific code, and the conversion table when the determination is absent determines the compressed digital signal. A digital signal having a means for expanding, a means for outputting the digital signal linked to the specific code when the discrimination is present, and a means for selecting these and expanding the compressed digital signal. Compression / expansion device.