JPS62236231A - Pcm transmitting system - Google Patents

Abstract

PURPOSE:To use the compressing transmission of difference data and to prevent a bad influence such as the dislocation of a direct current part due to a transmitting error by changing over and selecting the difference data and the linear data and supplying them to a quasi-instantaneous pressure companding CONSTITUTION:An A/D-converted original signal is given to a difference device 13 and a block maximum value detecting device 14, the difference device 13 successively obtains the difference value for each sample of the input and the detecting device 14 detects the maximum value (block maximum value) of the absolute value of a sample. A block difference maximum value detecting device 15 obtains a block differ ence maximum value from the output of the difference device 13. A discriminating device 16, based upon the output of detecting devices 14 and 15, discriminates whether or not the value is in an area C shown in Fig. (a). The discriminating device outputs discriminating data to a terminal 17, and controls a switch 18 so as to select the output of the difference device 13 or the output of a converting device 12 depending upon the fact that the output of the discriminating device 16 is '1' or '0.' A data compressing device 19 compresses them by the quasiinstantaneous pressure companding system and outputs the compressing data and the scale data. These data come to be NI-PCM data when the linear data are selected by the switch 18, and come to be NI-DPCM data when the difference.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention is based on PCM (pulse code mod
ulatlon (pulse code modulation) transmission system and its equipment, especially D PCM (dil'f'e
The present invention relates to a technique for improving DC component deviations during error correction in PCM transmission of a system that includes transmission of difference data between samples and includes data compression/expansion of transmission data, such as rent1al P CM ~difference PCM). In this case, "transmission" is not limited to transmission via wireless or wired communication lines, but simply means transmission in a broad sense, including transmission through modulation/demodulation systems and recording/reproduction systems. The term "transmission line" means a system including a communication/signal line, a modulation/demodulation system, a recording/reproduction system, etc., or a combination thereof.

[Conventional technology]

Simply convert analog signals into A/D (Analog-Digital)
Let us compare the case where PCM data that has just been converted, that is, linear data, and differential data, which is PCM data of difference values between samples, are respectively subjected to quasi-instantaneous companding. The PCM method for quasi-instantaneous companding of linear data is NI-
PCM (near instantaneous coB)
and lng P CM ~ quasi-instantaneous companding PCM). On the other hand, a PCM method for quasi-instantaneous companding of the difference data, that is, a differential quasi-instantaneous companding PCM (hereinafter referred to as rN I -DPCMJ for convenience of explanation) can be considered.

In Nl-PCM, the S/N is determined by the level of the original signal. On the other hand, in DPCM, the S/N is determined by the frequency and level of the original signal. Therefore, when the size of the original signal is smaller than the value corresponding to the transmission compression bit length, the S/N is the same in both N1-PCM and NI-DPCM, but when the original signal is If it is larger than the value corresponding to the compression bit length, N
In 1-PCM and N I-DPCM, S
/N is different. In other words, at low frequencies, even if the amplitude of the original signal is large, the value of the difference data is small, and therefore there is little loss of information when compressing the data, so N
The S/N of I-DPCM becomes good. However, this N
In I-DPCM, when there is a reduction in information due to compression and the above-mentioned data loss, a direct current deviation occurs. On the other hand, the value of linear data simply corresponds to the amplitude of the original signal. Therefore, if the amplitude of the original signal is small, the loss of information when compressing the data is small, and the S/N of N l-PCM
will be good. Furthermore, in Nl-PCM, since linear data is basically transmitted, no DC deviation occurs.

[Problem that the invention seeks to solve]

An object of the present invention is to realize PCM transmission with high efficiency and good S/H by using compressed transmission of differential data and substantially preventing adverse effects such as shifts in DC components caused by transmission errors. The purpose of this invention is to provide a PCM transmission system that can

[Means for solving problems]

The first invention according to the present invention is an invention of a method, which corresponds to linear data corresponding to the analog original signal and successive difference 1 values of the linear data based on an analog original signal or a signal obtained from the analog original signal. At the same time, the values of the linear data and the difference data are compared to determine whether the effective bit length of the absolute value of the linear data is larger than the predetermined compressed bit length, and the validity of the absolute value of the above-mentioned difference data is determined. It is substantially determined whether the bit length is less than or equal to the quantized bit length of the absolute value of the linear data, and according to the results of these determinations, the effective bit length of the absolute value of the linear data is less than the compressed bit length. If the difference data is large and the effective bit length of the absolute value of the difference data is less than or equal to the quantized bit length of the absolute value of the linear data, the difference data is used as the predetermined value. Obtain digital transmission data compressed according to the compression bit length,
The transmission data and discrimination information substantially corresponding to the above discrimination result are transmitted, the transmission data and discrimination information are received, the received transmission data is data-expanded, and the linear data is processed on the transmitting side based on the reception discrimination information. If selected, the expanded data is used as it is; if differential data is selected on the sending side, the expanded data is integrated,
P characterized in that it is used for conversion into an analog playback signal.
This is a CM transmission method.

Further, a second invention according to the present invention is an invention of a method that has the main part of the first invention as a main part and achieves the same object, and which provides an analog original signal or a signal obtained from the analog original signal. Based on this, linear data corresponding to the analog original signal and difference data corresponding to the sequential difference values of the linear data are obtained, and the values of these linear data and difference data are compared and determined to determine the effective bit length of the absolute value of the linear data. is greater than the predetermined compressed bit length,
and substantially determine whether the effective bit length of the absolute value of the above-mentioned difference data is larger than the effective bit length of the absolute value of the above-mentioned linear data, and according to the results of these judgments, the effective bit length of the absolute value of the above-mentioned linear data is determined. If the bit length is larger than the compressed bit length and the effective bit length of the absolute value of the difference data is larger than the effective bit length of the absolute value of the linear data, the linear data is used, otherwise the above Obtaining digital transmission data in which the differential data is compressed according to the predetermined compression bit length, transmitting the transmission data and discrimination information substantially corresponding to the discrimination result, and receiving the transmission data and discrimination information. to decompress the received transmission data,
In addition, based on the reception discrimination information, if linear data is selected on the transmitting side, the expanded data is used as is, and if differential data is selected on the transmitting side, the expanded data is integrated and analog playback is performed. This is a PCM transmission method that is characterized by being used for conversion into signals.

〔Example〕

Before specifically describing embodiments of the present invention, the principle of the present invention will first be explained.

The present invention is basically a method in which differential data and linear data are selectively selected and subjected to quasi-instantaneous companding during PCM transmission. When linear data is quasi-instantaneous companded, N1-P
Corresponds to CM, and when the difference data is quasi-instantaneous companded, N
Corresponds to I-DPCM. Therefore, in this method,
It can also be said that transmission is performed by switching between N1-PCM and N1-DPCM alternately.

In other words, N1-PCM and Nl-DPCM are switched according to the frequency domain and level domain of the original signal, and both S/N
In areas where there is a difference in S/N, the one with better S/N is selected, and in areas where the S/N between the two is substantially unchanged, Nl-P is selected.
CM and Nl-DPCM. In this case, N1-PCM has the advantage of being able to transmit a direct current component, and N1-DPCM has an advantage in terms of transmission efficiency.

Therefore, considering the transmission of DC components, we mainly use Nl-PCM.
Transmit by.

In other words, the effective bit length (effective bit length excluding the sign bit) of the absolute value of linear data obtained by sampling and quantizing the original signal is larger than the bit length of compressed data, and the frequency is lower than a predetermined value (sampling frequency If fs is the band below fS/12), Nl-D
PCM has a better S/N than Nl-PCM. And the absolute value of linear data is less than the bit length of compressed data.
When the effective bit length of is small, Nl-PCM
There is no difference in S/N between NI-DPCM and NI-DPCM.

Then, a region where the frequency is particularly high (f s / 6 to f s
/2 band), when the level of the original signal is high, the effective bit length of the absolute value of the difference data, which is the sequential difference of the linear data, may be larger than the quantization bit length of the linear data. In such a case, Nl-PCM, which has a smaller number of effective bits, definitely has a better S/N. In other high frequency areas (bands above fs/12), in principle, Nl-PCM and Nl-DPCM are used for S/
It can be assumed that there is no difference in N.

However, when compression errors due to data compression are strictly considered, N1-PCM has a substantially better S/N in the region where the effective bit length of the absolute value of the differential data exceeds the effective bit length of the absolute value of the linear data. .

Therefore, the following method can be considered as a first method.

In a region where the effective bit length of the absolute value of linear data is less than or equal to the bit length of compressed data, Nl-PCM is transmitted. In a region where the effective bit length of the absolute value of the linear data exceeds the bit length of the compressed data, N1-PCM is transmitted only in the region where the S/N is better, and the rest is transmitted by N1-DPCM. In other words, if the effective bit length of the absolute value of the linear data exceeds the bit length of the compressed data,
In a region where the frequency is not particularly high (band below fs/6), Nl-DPCM is used for transmission. Then, the effective bit length of the absolute value of the linear data exceeds the bit length of the compressed data and the frequency is particularly high (fs/6 to fs
/2 band), when the effective bit length of the absolute value of the differential data exceeds the quantization bit length of the linear data, Nl
-When transmitted using PCM and the effective bit length of the absolute value of the difference data is less than or equal to the quantization bit length of the linear data, N I-
Transmit using DPCM. This corresponds to the first aspect of the present invention.

In addition, the second
The following methods can be considered.

In a region where the effective bit length of the absolute value of linear data is less than or equal to the bit length of compressed data, Nl-PCM is transmitted. In a region where the effective bit length of the absolute value of the linear data exceeds the bit length of the compressed data and the frequency is not particularly high (band below fs/6), N1-DPCM is used for transmission. and,
A region where the effective bit length of the absolute value of linear data exceeds the bit length of compressed data and where the frequency is particularly high (f s /
6 to fs/2), if the effective bit length of the absolute value of the differential data exceeds the effective bit length of the absolute value of the linear data, it is transmitted in N1-PCM, and the effective bit length of the absolute value of the differential data is If the length is less than or equal to the effective bit length of the absolute value of linear data, it is transmitted using Nl-DPCM. This corresponds to the second aspect of the present invention.

For example, when PCM data is subjected to quasi-instantaneous companding processing that compresses 18-bit original sample data to 8 bits during transmission, if the original sample data to be quasi-instantaneously companded is linear data, N1- When the data is PCM data and the original sample data to be subjected to quasi-instantaneous companding is differential data, it is Nl-DPCM data.

FIG. 3 shows the relationship between signal level (S/N) and frequency.

Characteristic A shown by a solid line is a characteristic of NI-DPCM, and characteristic B shown by a broken line is a characteristic of Nl-PCM. The dynamic range of both of these is approximately 98 dB, which is similar to 16-bit linear PCM.

When the frequency of the original signal is in the range of fs/2 to fs/6, the effective bit length of the absolute value of the signal frequency, level, and difference data (hereinafter, "effective bit length of absolute value" needs to be particularly distinguished) The relationship between the signal level and the effective bit length of linear data (the effective bit length of linear data does not depend on the signal frequency) is shown in the fourth section, respectively.
Shown in Figures (a)p and (b). As can be seen from Fig. 4 (a) and (b), the frequency of the original signal is f s /
in the range of 2 to f s / 6 and at the maximum level (Od
In the case of B), the differential data value is 17 bits, which is larger than the quantization bit length of linear data, which is 16 bits. Therefore, if the bit lengths of compressed data are the same, Nl-DPCM should have a larger error due to compression, and the S/N will be worse than Nl-PCM.

In the quasi-instantaneous companding, a value corresponding to the number of cut-off bits when the differential data or linear data is compressed to 8 bits consisting of the upper effective bits is transmitted together with the compressed data as a scale value. Since this scale value depends on the effective bit length, based on Figures 4(a) and (b), the scale value, frequency, and level when differential data or linear data is quasi-instantaneous companding are calculated. The relationship is schematically shown in a part of FIG. The region S shown in FIG. 3 corresponds to the region where the previously mentioned difference data value is 17 bits, which is larger than 16 bits, which is the quantization bit length of the linear data. Regions T and S are Nl-DP
This is a region where the scale value of Nl-PCM is smaller than the scale value of CM, and for this region, N1-PCM
Since the number of truncated bits is smaller for M, N1-
PCM has better S/N.

In addition, when the frequency of the original signal is in the range of fs/6 to fs/12, as is clear from Fig. 3, NI-DPCM and N1
- PCM S/N is equal. This means that in this area, the effective bit lengths of differential data and linear data are the same.

In the region where the frequency of the original signal is lower than fs/12, NI-DPCM
S/N improves by dB. In this frequency domain, when the level of the original signal is higher than the level corresponding to the bit length of compressed data, NI-DPCM has a better S/N.
can be obtained. In this frequency domain, when the level of the original signal is equal to or lower than the level corresponding to the bit length of compressed data, the S/N is the same in both NI-DPCM and Nl-PCM.

Therefore, in the region where the level of the original signal is below the level corresponding to the bit length of the compressed data, that is, in the region below the broken line B in FIG. 3, N1-PCM data is transmitted. In the area where the level of the original signal exceeds the level corresponding to the bit length of the compressed data, that is, the area above the broken line B in FIG. 3, N1-PCM provides better transmission than N1-D PCM. Nl-DPCM data is transmitted and Nl-PC
If M can perform better transmission than Nl-DPCM, N1-PCM data is transmitted.

In other words, in the area where the level of the original signal exceeds the level corresponding to the bit length of the compressed data, that is, the area above the broken line B in FIG. 3, the area where the difference data value is larger than the quantized bit length of the linear data. Then N1-PCM
Data is transmitted, and Nl-DPCM data is transmitted in a region where the differential data value is less than or equal to the quantization bit length of the linear data (corresponding to the first invention).

That is, for the region C within the thick line shown in FIG.
-DPCM data, N1-P for other areas
Transmission is performed using CM data.

In addition, among areas where the level of the original signal exceeds the level corresponding to the bit length of compressed data, N I-P CM data is transmitted in areas where the effective bit length of differential data is greater than the effective bit length of linear data. However, in a region where the effective bit length of the differential data is less than or equal to the effective bit length of the linear data, NI-DPCM data is transmitted (corresponding to the second invention).

In this case, in the area above the broken line B in FIG. 3, it is determined whether to transmit Nl-PCM data or Nl-DPCM data.
- It may be performed based on the scale value of the DPCM data. In other words, the scale value of NI-DPCM data is N
When exceeding the scale value of l-PCM data, N1-
When transmitting PCM and the scale value of NI-DPCM data is less than or equal to the scale value of Nl-PCM data, N
l-DPCM is transmitted.

That is, for the shaded area shown in FIG.
Regarding PCM data and other areas・□ is N l
-Transmission is performed using PCM data.

In this way, good transmission characteristics can always be obtained.

The structure of a first embodiment of the present invention based on such a principle is shown in FIGS. 1 and 2. This first embodiment is an embodiment to which the above first invention is applied, and FIG. 1 shows a transmitting device and FIG. 2 shows a receiving device.

In the transmitter shown in FIG. 1, an analog original signal is input to the input terminal 11, and this analog signal is
The D converter 12 samples and digitizes the signal at a predetermined frequency. The output of this A/D converter 12 is given to a differentiator 13 and a block maximum value detector 14. The differentiator 13 sequentially obtains difference values for each sample of input original sample data. Block maximum value detector 1
4 detects the maximum absolute value of the samples in the block, that is, the block maximum value, for each data block consisting of a plurality of samples of original sample data. The output of the differentiator 13 is given to a block difference maximum value detector 15.

The block difference maximum value detector 15c determines, for each data block, the maximum absolute value of the difference data in that block, that is, the block difference maximum value, from the difference data output from the difference unit 13. The block maximum value outputted from the block maximum value detector 14 and the block difference maximum value outputted from the block difference maximum value detector 15 are given to the discriminator 1B. The discriminator 1B determines the bit length of the maximum value of linear data based on the block maximum value, which is the maximum value of linear data, and the block difference maximum value, which is the maximum value of differential data, in each data block. By determining that the compressed bit length (8 bits) is exceeded, and that the effective bit length of the maximum value of the difference data is less than or equal to the predetermined quantization bit length (16 bits), Determine the signal level and frequency of a signal.

That is, the discrimination performed by the discriminator 1G is to determine whether or not the original signal is within the region C shown in FIG. Specifically, the discriminator 1B can be configured as shown in FIG. 5, in which the block maximum value is compared with the first reference data ("11111111">) by the comparison means 18a, and the block difference maximum value is compared. The means 16b stores the second reference data (“
1111111111111111'').Then, the AND means 18c compares these two comparison means 18.
By calculating the logical product of the comparison results of a and 18b, it is determined that the block maximum value exceeds the first reference data ("11111111°"), and the block difference maximum value exceeds the second reference data ("
1l1111H1llllllll'), outputs "1", and otherwise outputs "0#".

Then, the discriminator 18 outputs discrimination data indicating the discrimination result to the output terminal 17, and controls the switching means 18 according to the discrimination result.

The switching means 18 switches and selects the output of the A/D converter 12 and the output of the differentiator 13 to select the output of the data compressor 1.
Give to 9. The selection of the switching means 18 is determined by the discriminator 1
Select the output of the differentiator 13 when the output of B is "1",
'0' selects the output of the A/D converter 12. In this case, the data compressor 19 encodes the input data using a quasi-instantaneous companding method, that is, compresses the data. Compressed data and scale data are obtained and output to output terminals 20 and 21, respectively.In this quasi-instantaneous compression encoding, when the effective bit length of the input data exceeds the bit length of the compressed data, the upper effective bits are In addition to outputting compressed data, it also outputs scale data indicating the bit position from which the compressed data was extracted.In this case, data compression processing is performed for each block, so the maximum (absolute) value data in the block is When the effective bit length exceeds the bit length of compressed data, for all data in the block, the data at the bit position corresponding to the upper effective bit of the maximum (absolute) value data is set as compressed data, and the bit position is Let the data indicating this be the scale data for the block.

These compressed data and scale data are Nl when linear data is selected by the switching means 18.
-N when PCM data or differential data is selected
This becomes l-DPCM data. The compressed data, scale data and discrimination data outputted from the output terminals 20, 21 and 17 are appropriately multiplexed by a time division multiplexing means (not shown) and sent to the transmission system. Of course, in this case, the signals may be sent to separate transmission paths without being multiplexed.

That is, in this transmitter, a plurality of pieces of sampling data are treated as one block, and the maximum (absolute) value of linear data and the maximum (absolute) value of difference data in the block are detected by the block maximum value detector 14 and the block difference detector 14, respectively. The maximum value detector 15 detects these, and the discriminator 16 discriminates them. As mentioned above, this discriminator 16 determines whether the original signal is
It is determined whether it is within area C in the figure. The switching means 18 is switched according to the result of this determination, and if it is within the region C, the difference data which is the output of the differentiator 13 is selected, and if it is not, the linear data which is the output of the A/D converter 12 is selected. do. The selected data is compressed by the data compressor 19 according to the quasi-instantaneous companding method. This data compressor 19 obtains compressed data and scale data indicating a compression scale factor for each block, and these compressed data and scale data are transmitted. This transmission data further includes whether it was sent in N1-PCM or Nl-
In order to enable the receiving side to determine whether the data was sent by DPCM, determination data that substantially indicates the determination result of the discriminator 1B is added.

In the receiving device shown in FIG.
Compressed data, scale data, and discrimination data transmitted and received from the transmitting device are input to 2 and 33, respectively. When these transmission data are multiplexed and transmitted, they are separated by appropriate separation means (not shown) and then input to these input terminals 31, 32 and 33. When the transmission data is transmitted via separate transmission paths without being multiplexed, the compressed data, scale data, and discrimination data are input as they are to the respective input terminals 31, 32, and 33, respectively. The compressed data and scale data are provided to a data decompressor 34 that follows a quasi-instantaneous companding method. The data decompressor 34 is
In this case, the compressed data is decoded using a quasi-instantaneous companding method, that is, data expanded, using the scale data, and the compressed data is bit-shifted according to the scale data and output. The output of this data expander 34 is given to an integrator 35 for demodulating the difference data. Also,
The discrimination data applied to the input terminal 33 is applied to the switching means 36. The switching means 36 is configured as a changeover switch that responds to the discrimination data, and selects either the output of the data expander 34 or the output of the integrator 35. The data selected by this switching means 3G is given to a D/A (digital-to-analog) converter 37 and converted into an analog value. D/
The analog signal output from the A converter 37 is output via the output terminal 38 and is provided for reproduction.

That is, compressed data and scale data, which are transmission data, are led to the data decompressor 34 and are decompressed. The switching means 36 is controlled by the discrimination data, which is also transmission data, and the integrator 35 is controlled depending on whether linear data or differential data is selected on the transmitting side.
It is possible to switch whether to pass or not. In other words, when linear data is selected, N1-PCM transmission is performed, so the output of the data expander 34 is sent directly to the D/A converter 37.
When differential data is selected, the integrator 35
After being integrated, the signal is converted into an analog value by a D/A converter 37 to obtain an analog reproduction signal.

In this case, for signals such as general music signals and speech, the time required to transmit Nl-PCM is relatively long, and Nl-DPCM transmission continues for a long time without intervening N1-PCM transmission. There is little to do. Even if an uncorrectable error occurs due to missing data during signal transmission and is corrected, the problem of DC component deviation will not occur during N1-PCM transmission. Also, N I-
Even if a deviation occurs in the DC component due to an error or data compression during DrCM transmission, the deviation is substantially corrected by performing Nl-PCM transmission while the deviation is very small. Therefore, in practice: There is no qualitatively problematic deviation in the direct current component.

In this way, the level and frequency of the original signal are determined by determining the magnitude of the absolute value of the linear data and the difference data obtained from the original signal, and N1-PCM transmission and Nl
- By transmitting in combination with DPCM transmission, it is possible to widen the region with good S/H in highly efficient PCM transmission.

Moreover, in the low level range and high frequency high level range, N1-PCM transmission is used, so there is no deviation of the DC component due to transmission errors. appears with high frequency in general music signals, speech signals, etc., and as a result, it is possible to effectively reduce DC deviations in overall transmission.

In addition, in this case, N1-PCM is used only when the bit length of linear data is smaller than the compression bit length and when the data length of differential data is larger than the quantization bit length. Scale value is “O”
and “8”, and no other values are accepted. therefore,
The types (possible numbers) of scale values and discrimination information to be transmitted with compressed data are N1-PCM scale values "0" and "8'" and NI-DPCM (7) scale values "O~8". Since there are only 10 types of ``,'' information to be transmitted together with the compressed data can be sent in 4 bits.

Therefore, it is possible to perform extremely highly efficient PCM transmission with a very small amount of transmitted information, to ensure a high S/N ratio, and to significantly reduce the deviation of the DC component.

Next, FIG. 6 shows the configuration of a transmitting apparatus according to a second embodiment of the present invention to which the above-mentioned second invention is applied.

The receiving device in this second embodiment may have substantially the same configuration as that shown in FIG. 2, so a description thereof will be omitted here.

In the transmitting device shown in FIG. 6, only the discriminator 22 differs from the discriminator 14 in the configuration of FIG. 1, and the rest is substantially the same as the configuration of FIG. 1. Therefore, detailed description of the configuration similar to that in FIG. 1 will be omitted.

In this case, the block maximum value output from the block maximum value detector 14 and the block difference maximum value output from the block difference maximum value detector I5 are provided to the discriminator 22. The discriminator 22 determines the bit length of the maximum value of linear data based on the block maximum value, which is the maximum value of linear data, and the block difference maximum value, which is the maximum value of differential data, in each data block. By determining that the compressed bit length (8 bits) of Determine the signal level and frequency of.

That is, the discrimination performed by the discriminator 22 is to determine whether or not the original signal falls within the range shown in FIG. Specifically, the discriminator 22 can be configured as shown in FIG. 7, in which the block maximum value is compared with the (first) reference data ("11111111") by the comparing means 22a, and the block difference maximum value is determined. is compared with the block maximum value by the comparing means 22b. Then, by the logical product means 22c, these two comparison means 22
By calculating the logical product of the comparison results of a and 22b, the block maximum value exceeds the reference data (“11111111”),
In addition, when the block difference maximum value is less than or equal to the block maximum value, "1" is output, and otherwise, "0" is output.

Then, the discriminator 22 outputs discrimination data indicating the discrimination result to the output terminal 17, and controls the switching means 18 according to the discrimination result.

The switching means 1B switches and selects the output of the A/D converter 12 and the output of the differentiator 13 to select the output of the data compressor 1.
Give to 9. The selection of the switching means 18 is determined by the discriminator 2
When the output of 2 is "1", select the output of the differentiator 13,
When the angle is 0°, the output of the A/D converter 12 is selected.The data selected by the switching means 18 is encoded, that is, data compressed, by a quasi-instantaneous companding method in the data compressor 19, and is sent to the output terminal 20.21. Output.

That is, in this transmitter, a plurality of pieces of sampling data are treated as one block, and the maximum (absolute) value of linear data and the maximum (absolute) value of difference data in the block are detected by the block maximum value detector 14 and the block difference detector 14, respectively. The maximum value detector 15 detects these, and the discriminator 22 discriminates them. This discriminator 22 detects that the original signal is the third
It is determined whether or not it is within the range shown in the figure, and the switching means 18 is switched according to the result of this determination, and if it is within the range, the difference data that is the output of the subtractor 13 is selected, and if it is not, it is selected. Select the linear data that is the output of the A/D converter 12. The selected data is compressed by the data compressor 19 according to the quasi-instantaneous companding method. This data compressor 19 obtains compressed data and scale data indicating a compression scale factor for each block, and these compressed data and scale data are transmitted. This transmission data also includes whether it was sent in Nl-PCM or Nl-D.
In order to enable the receiving side to determine whether the data was sent using PCM, determination data that substantially indicates the determination result of the discriminator 1B is added.

In this case, in general signals, N1-PC
The time transmitted by M is relatively long, and Nl-DPC
The fact that M transmission almost never continues for a long time without intervening N1-PCM transmission, and that there is virtually no problem in the DC component at all, is almost the same as in the case of the first embodiment described above. The same is true. In this way, the level and frequency of the original signal are determined by determining the absolute values of the linear data and difference data obtained from the original signal, and N1-PCM transmission and N I-DPCM transmission are performed according to the determination results. By transmitting in combination, S
/N can be expanded. Moreover, in low level areas and areas where Nl-DPCM transmission has more truncation errors due to compression, Nl-PCM transmission is used, which improves the S/N ratio and at the same time almost eliminates the influence of DC component shifts due to transmission errors. .

Note that it is also possible to determine whether or not the original signal is within the area shown in FIG. 3 by directly comparing the scale information of N1-PCM and the scale information of N1-DPCM. In this case, the linear data output from the A/D converter 12 and the difference data output from the subtractor 13 are data compressed, and Nl-PCM and Nl-DPC are
Obtain M data (compressed data and scale value), find the maximum value of each scale data in the data block, and provide both to the discriminator. The discriminator uses a scale value of N1-PCM that is not 0° and a scale value of N1-PCM that is N
NI when not larger than the scale value of l-DPCM
- Select DPCM data and provide for transmission.

Note that the present invention is not limited to the embodiments described above and shown in the drawings, and can be implemented with various modifications within the scope of the gist thereof.

〔Effect of the invention〕

According to the present invention, compressed transmission of sequential differential data is effectively used, problems such as DC component deviation are substantially solved, and PCM transmission with high efficiency and good S/N ratio is realized. It is possible to provide a PCM transmission method that obtains the following information.

[Brief explanation of drawings]

1 and 2 are block diagrams showing the configurations of a transmitting device and a receiving device, respectively, in a first embodiment of the present invention, FIGS. 3 and 4 are diagrams for explaining the principle of the present invention, and FIG. FIG. 5 is a block diagram schematically showing the configuration of the main parts in the first embodiment of the present invention, FIG. 6 is a block diagram showing the configuration of the transmitting device in the second embodiment of the present invention, and FIG. FIG. 2 is a block diagram schematically showing the configuration of main parts in a second embodiment of the present invention. 12...A/D (analog-digital) converter, ■3...Differentiator, 14...Block maximum value detector, 15...Block difference maximum value detector, 18.22
... Discriminator, 18, 38 ... Switching means, 1
9... Data compressor, 34... Data decompressor, 35
...Integrator, 37...D/A (digital-
analog) converter. Applicant's representative Patent attorney Takehiko Suzue No. 5! ! ! ! 1 Figure 1'll-PCM Nuguere supplementary fishing maximum supplement 9-s/e heg-

Claims (6)

[Claims] (1) Based on the analog original signal or a signal obtained from the analog original signal, obtain linear data corresponding to the analog original signal and difference data corresponding to successive difference values of the linear data, and obtain these linear data and difference data. Compare and determine whether the effective bit length of the absolute value of the linear data is larger than the predetermined compressed bit length.
and substantially determine whether the effective bit length of the absolute value of the above-mentioned difference data is less than or equal to the quantized bit length of the absolute value of the above-mentioned linear data, and according to the results of these judgments, the effective bit length of the absolute value of the above-mentioned linear data If the effective bit length is larger than the compressed bit length and the effective bit length of the absolute value of the difference data is less than or equal to the quantized bit length of the absolute value of the linear data, the difference data is Digital transmission data in which the linear data is compressed according to the predetermined compression bit length is obtained, the transmission data and discrimination information substantially corresponding to the discrimination result are transmitted, and the transmission data and discrimination information are is received, the received transmission data is data decompressed, and based on the reception discrimination information, if linear data is selected on the transmitting side, the decompressed data is used as is, and if differential data is selected on the transmitting side, the decompressed data is used as is. A PCM transmission method characterized in that the expanded data is integrated and converted into an analog reproduction signal. (2) The PCM transmission method according to claim 1, wherein the determination is made for each data block consisting of a plurality of data, and the maximum value data within the data block. (3) Obtain linear data corresponding to the analog original signal and difference data corresponding to successive difference values of the linear data based on the analog original signal or a signal obtained from the analog original signal, and obtain these linear data and difference data. Compare and determine whether the effective bit length of the absolute value of the linear data is larger than the predetermined compressed bit length.
and substantially determine whether the effective bit length of the absolute value of the above-mentioned difference data is larger than the effective bit length of the absolute value of the above-mentioned linear data, and according to the results of these judgments, the effective bit length of the absolute value of the above-mentioned linear data is determined. If the bit length is larger than the compressed bit length and the effective bit length of the absolute value of the difference data is larger than the effective bit length of the absolute value of the linear data, the linear data is used, otherwise the above Obtaining digital transmission data in which the differential data is compressed according to the predetermined compression bit length, transmitting the transmission data and discrimination information substantially corresponding to the discrimination result, and receiving the transmission data and discrimination information. to decompress the received transmission data,
In addition, based on the reception discrimination information, if linear data is selected on the transmitting side, the expanded data is used as is, and if differential data is selected on the transmitting side, the expanded data is integrated and analog playback is performed. A PCM transmission method characterized by being used for conversion into signals. (4) In the PCM transmission system according to claim 3, the discrimination and data compression are performed for each data block consisting of a plurality of data based on the maximum value data in the data block. PCM transmission method. (5) PC according to claim 3 or 4
The PCM transmission method is characterized in that data compression is quasi-instantaneous compression in which compressed data and compression scale information are substantially transmitted. (6) In the PCM transmission system according to claim 5, the comparison between the effective bit length of the absolute value of the differential data and the effective bit length of the linear data is performed using a compression scale value. Transmission method.