JPH08223039A - Linear error reduction device and linear error detection device for analog-digital converter - Google Patents

Abstract

PURPOSE: To provide a linear error reduction and a linear error detection device which can reduce the distortions of digital audio signals caused by the linear error of an A/D converter. CONSTITUTION: The digital audio signals are inputted to a histogram production means 3 via a switch means 1, and the emerging frequency is detected for every code of the digital audio signals. Then a normalization means 4 extracts a linear error as a special feature of the emerging frequency by dividing the emerging frequency of every code by the average emerging frequency of every code and the codes having their quantization levels approximate to every code. An error bit detection means 5 detects an error bit based on the output of the means 4. Then a digital signal processing means 6 inputs the digital audio signals via the means 1 and outputs the digital audio signals that have distortions caused by the linear error of an A/D converter and then reduced based on the output of the means 5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a linearity error reducing device for reducing the distortion of a digital audio signal having a distortion due to a linearity error generated during A / D conversion, and an error detection in this linearity error reducing device. The present invention relates to a suitable linearity error detection device.

[0002]

2. Description of the Related Art Conventionally, the linearity error generated in an A / D converter has been reduced by adjusting the internal resistance value of the A / D converter or adjusting the bit weight by an external circuit. . In the conventional method, the linearity error can be reduced at the time of A / D conversion. However, the distortion cannot be reduced if it is once stored in a state where the distortion is caused by the linearity error at the time of A / D conversion. Many music software programs that have been digitally recorded using an A / D converter in the past are accompanied by distortion due to a considerable linearity error due to poor conversion accuracy of the A / D converter in the past. But,
It was impossible to reduce this by the conventional method.

On the other hand, as a device for detecting a linearity error generated in an A / D converter, a measuring device using a sine wave histogram method has been conventionally known. However, with this measuring device, a pure sine wave without distortion is
The linearity error is measured by inputting it to the / D converter and obtaining a histogram. Therefore, when a signal which is not a pure sine wave without distortion, such as a digital audio signal, is used as the signal source, such a test apparatus cannot detect the linearity error itself.

[0004]

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and is a linearity error for reducing the distortion of a digital audio signal having distortion based on the linearity error at the time of A / D conversion. It is an object of the present invention to provide a reduction device. A linearity error detection device suitable for use in this linearity error reduction device, capable of detecting linearity error even when a pure sine wave without distortion is not used as a signal source. The purpose is to provide a device.

[0005]

According to a first aspect of the present invention, in a linearity error reducing device for an A / D converter, at least a digital audio signal converted into a code by the A / D converter is provided. An appearance frequency detecting means for inputting a part and detecting the appearance frequency of the code,
Linearity error extraction means for inputting the output of the appearance frequency detection means and extracting the linearity error of the A / D converter, and inputting the digital audio signal for outputting the A / D based on the output of the linearity error extraction means. It is characterized in that it has an error reducing means for reducing the distortion due to the linearity error of the converter.

In the invention described in claim 2, the A / D
In a linearity error detection device for a converter, an appearance frequency detecting means for inputting at least a part of a digital audio signal converted into a code by an A / D converter and detecting the appearance frequency of the code, and an output of the appearance frequency detecting means. The linearity error of the A / D converter is extracted as a feature of the appearance frequency by inputting and dividing the appearance frequency for each code by the average appearance frequency in each code and the code whose quantization level is in the vicinity of each code. It is characterized in that it has a linearity error extracting means.

[0007]

According to the present invention, in the invention described in claim 1, the appearance frequency is detected by inputting at least a part of the digital audio signal converted into the code by the A / D converter and detecting the appearance frequency of the code. Means, a linearity error extracting means for inputting an output of the appearance frequency detecting means and extracting a linearity error of the A / D converter, and an A based on an output of the linearity error extracting means for inputting the digital audio signal. Even if the characteristics of the A / D converter used are unknown, the A / D converter has an error reducing means for reducing the distortion due to the linearity error of the A / D converter.
The distortion due to the linearity error of the D converter can be reduced.

In the invention described in claim 2, the A / D
An appearance frequency detecting means for inputting at least a part of the digital audio signal converted into the code by the converter and detecting the appearance frequency of the code, and inputting the output of the appearance frequency detecting means, the appearance frequency of each code And its quantization level has a linearity error extracting means for extracting the linearity error of the A / D converter as a feature of the appearance frequency by dividing it by the average appearance frequency in the codes near the respective codes. Used A /
The linearity error of this A / D converter can be detected without the D converter itself.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram for explaining the outline of an embodiment of the present invention. In the figure, 1 is a switching means, 2 is a memory, 3 is a histogram forming means, 4 is a normalizing means, 5 is an error bit detecting means, and 6 is a digital signal processing means. As a first step, a digital audio signal is read from a digital audio tape recorder (DAT), a compact disc (CD) or the like and input to the switching means 1 via a digital audio interface or the like. This digital audio signal is an analog audio signal converted and recorded by an A / D converter before being recorded on a compact disc or the like. The read digital audio signal may be a part of that recorded on a digital audio tape recorder or the like. The switching means 1 outputs the digital audio signal to a memory 2 such as a hard disk, and the memory 2 stores the digital audio signal.

The histogram forming means 3 reads the digital audio signal once stored in the memory 2, detects the appearance frequency of each code representing a quantized amplitude level of a plurality of samples, and creates a histogram. To do. The normalizing means 4 is the histogram forming means 3.
Is input to extract the linearity error of the A / D converter as a feature of the appearance frequency and output as a normalized histogram. The error bit detection means 5 inputs the normalized histogram from the normalization means 4, detects the error bit from the characteristic of the appearance frequency due to the linearity error of the A / D converter, and outputs the detection result to the digital signal processing means 6 Output to.

In the second step, the digital audio signal is read again from the digital audio tape recorder or the like and input to the switching means 1 via the digital audio interface or the like. The switching means 1 outputs the digital audio signal to the digital signal processing means 6. The digital signal processing means 6 inputs the digital audio signal, and the error bit detecting means 5
On the basis of the output of, the audio signal converted into an erroneous code due to the linearity error of the A / D converter is converted into a value close to the original correct level, thereby performing a process of reducing the linearity error in real time. The digital audio signal in which the distortion due to the linearity error is corrected is output. The digital signal processing means 6 is realized by a digital signal processing processor (DSP) or the like.

In the above description, in the second step, the digital audio signal is read again from the digital audio tape recorder or the like, but all the digital audio signals to be corrected for the distortion due to the linearity error are stored in the memory 2. When the data is stored in the memory 1, the output of the memory 2 may be directly connected to the input of the digital signal processing means 6 via the switching means 1. When a large-capacity storage device composed of a semiconductor memory such as a RAM is used as the memory 2, even with such a configuration, the linearity error is corrected in real time after receiving the output from the error bit detection means 5. The output digital audio signal can be output.

When it is not necessary to output a digital audio signal whose linearity error has been corrected in real time, for example, when the output is stored in another storage device, a hard disk having a low data transfer rate is used as the memory 2. Alternatively, the output of the memory 2 may be connected to the input of the digital signal processing means 6 via the switching means 1 or directly.

When a digital audio signal is input from a storage device composed of a semiconductor memory such as a ROM or RAM and a digital audio signal whose linearity error is corrected is output, the storage device itself is the memory 2. By controlling the storage device so as to achieve the function (1), the output of this storage device can be directly connected to the inputs of the histogram forming means 3 and the digital signal processing means 6.

The outline of the embodiment of the present invention is as described above, and the function of each block shown in FIG. 1 will be described in detail below.

FIG. 2 is an example of a histogram of an actual digital audio signal. In the figure, the horizontal axis represents the code of the digital audio signal in decimal, and the vertical axis represents the frequency of appearance for each code. The code of the digital audio signal represents the amplitude level quantized by the A / D converter. This histogram has a high appearance frequency when the code is 0, and the appearance frequency gradually becomes gentle as the code increases in the positive direction and decreases as the code decreases in the negative direction. The frequency of appearance is substantially bilaterally symmetrical in the positive and negative directions centering around the case where the code is 0. The features of such a histogram are unique to each input audio signal. In the illustrated histogram, there are also codes in which the frequency of occurrence sharply increases and codes in which the frequency of occurrence sharply decreases as compared with the adjacent codes. This peculiar appearance frequency is caused by the linearity error of the A / D converter.

The histogram forming means 3 shown in FIG.
A device similar to the conventional sine wave histogram measuring device can be used. The digital audio signal is read out from the memory 2, the histogram memory is accessed by using the code itself representing the quantized amplitude level as an address, and the data at this address is read out.
Is added and the added value is written to the same address. By performing this operation on multiple samples,
The cumulative frequency for each code, that is, the frequency of appearance is stored in the histogram memory. The plurality of samples do not have to be all the digital audio signals to be corrected for the linearity error, but may be some samples converted by the same A / D converter. However, the accuracy of the histogram improves as the number of samples increases. As an example, a good histogram distribution can be obtained by using a sample for a period of 2,500,000 samples in which relatively high-level portions are consecutive. The histogram conversion can be realized by a digital counter circuit or software.

The normalizing means 4 shown in FIG. 1 receives the output of the histogram forming means 3 and compares it with the frequency of appearance peculiar to the input audio signal to determine the linearity error of the A / D converter. Is extracted as a feature of. The average frequency of appearance of each code and the code whose quantization level is in the vicinity of each code represents the frequency of occurrence peculiar to the input audio signal. Therefore, as an example, the linearity error of the A / D converter can be extracted as a feature of the appearance frequency by dividing the appearance frequency for each code by the average appearance frequency. Specifically, all codes are divided into a plurality of blocks with a plurality of codes having adjacent quantized amplitude levels, for example, four codes as one unit. The average frequency of appearance for each block is obtained from the frequency of occurrence of each code forming this block.

A value obtained by dividing the appearance frequency of each code forming a block by the average appearance frequency of each block is defined as a normalized appearance frequency A of each code. That is,
(Frequency of occurrence of each code divided into 4) ÷
(Average of appearance frequency of four codes) = A. The histogram of the normalized frequency of appearance is A / D
The feature of the appearance frequency due to the linearity error of the converter is extracted. The normalized appearance frequency A is obtained by almost removing the characteristic of the appearance frequency peculiar to the audio signal. If the A / D converter has no linearity error,
The normalized appearance frequency A is 1 for all codes. As the average frequency of occurrence, a moving average of each frequency of occurrence of each code and a predetermined number of codes whose quantization levels are near the code may be used.

As a special case, a digital audio signal read from a digital audio tape recorder or the like may include a pure sine wave signal or a digitized musical tone signal.
When such a signal is present, the normalized frequency of appearance can be obtained by the same method as the calculation of the differential linearity error in the conventional sine wave histogram measuring apparatus. A pure sine wave histogram without distortion is theoretically obtained, and it is small when the amplitude level is 0, gradually becomes bilaterally steep in positive and negative directions, and becomes maximum at the positive and negative peak levels. It will be.
This theoretically obtained histogram represents the appearance frequency peculiar to the input audio signal. Input a digital audio signal of a pure sine wave signal without distortion or a tone signal close to this, and divide the appearance frequency of each code by the appearance frequency of the same chord obtained from the theoretical value of the sine wave histogram. Can obtain the normalized appearance frequency. The value obtained by subtracting 1 from this value is the differential linearity error.

The error bit detecting means 5 receives the normalized histogram output from the normalizing means 4 and detects an error bit. The normalized histogram includes codes with extremely high frequency of occurrence and codes with extremely low frequency of occurrence, and the difference in the quantization level between the codes with high frequency and the frequency with low frequency is the weight of the error bit in the A / D converter. In other words, it represents the digit position of a bit.
The cause of the linearity error depends on the method of the A / D converter when the audio signal is digitized. Here, a weight resistance type D / A converter or R-2
It is premised on a successive approximation type A / D converter using a binary weight current source type D / A converter having an R ladder as an internal A / D converter. That is, assuming that the weight is used to set the amplitude level as the threshold,
Focusing on the influence of a certain weight on multiple quantized amplitude levels, conversely, detect the weight that caused the linearity error from multiple quantized amplitude levels, that is, the frequency of occurrence of the code. Is to do.

For example, when the interval between codes having extremely high occurrence frequencies is 2, that is, when the interval of the quantization level is 2 ¹ times the quantization level of the least significant bit, an error is generated in the least significant bit. I know there is. If the interval is 2 ³ , that is, the quantization level interval is 2 ³ times the quantization level of the least significant bit, there is an error in the bit at the third digit position from the least significant bit. Recognize. When the quantization level interval of a plurality of codes having an extremely high frequency of occurrence is 2 ⁿ , it can be seen that there is an error in the bit at the digit position of the n-th bit from the least significant bit. A threshold is required to determine that the frequency of appearance is extremely high. For example, a normalized appearance frequency A of 1.2 or more can be determined to have an extremely high appearance frequency. This threshold is 1.5
It may be about. On the contrary, it is possible to detect the digit position of a bit having an error also from the interval between codes having extremely low occurrence frequencies. Alternatively, both can be detected in combination.

FIG. 3 is a flow chart for explaining the processing steps for realizing the function of the error bit detecting means. S
At 11, the normalized histogram is input from the normalizing means 4. In S12, the initial value of the value of the variable n is set to 1 from the quantization bit number of the digital audio signal.
Is subtracted. In S13, the normalized appearance frequency of each code is extracted for each quantization level 2 ⁿ of each code. In S14, it is determined whether or not all of the extracted normalized appearance frequencies are 1.2 or more. If 1.2 or more, the process proceeds to S15, where 1.2
If less, the process proceeds to S16. In S15,
Since it is known that there is an error at the nth bit from the least significant bit, the value of n is stored in the error bit memory. In S16, 1 is subtracted from the value of n to update the value of n. S1
If the value of n is 0 in 7, the process proceeds to S18,
If the value of n is not 0, the process returns to S13. In S18, the error bit stored in the error bit memory is transferred to the digital signal processing means 6 and the process is ended.

The digital signal processing means 6 receives the digital audio signal as input and performs processing for reducing the linearity error in real time based on the output of the error bit detecting means 5. Distortion due to the linearity error of the A / D converter is reduced by converting a part of the plurality of samples having a level having a high appearance frequency due to an error into a level which is supposed to be an original correct level.

FIG. 4 is a flow chart showing the procedure for correcting the linearity error. In S21, the code of each sample of the digital audio signal to be corrected is input in real time. In S22, it is determined whether the value at the digit position of the error bit is 1 for each code of one sample. A code whose value at the digit position of the error bit is 1 is a code whose quantization level is affected by the error bit. If the value at the digit position of the error bit is 1, the process proceeds to S23, where 1
If not, the process proceeds to S24. In S23, the correct quantization level of the sample, that is,
The code is estimated and the process proceeds to S24. The code of the sample input in S24 or S23
The code estimated in 1. is written to an internal file and then output to the outside. At this time, the position on the time axis,
That is, the data indicating which sample the sample is from the beginning is also stored in the file together with the code.

Let LSB be the quantization level of the least significant bit, a and b be positive integers, and N be 0 or a positive integer. When the resistance value inside the successive approximation type A / D converter has an error of + b · LSB at the a-th bit from the lower order, the frequency of occurrence of a code having a quantization level of 2 ^a · N + 2 ^a-1 −1 is b + 1. And the quantization level is 2 ^a
The appearance frequency of each code between N + 2 ^a −1 and 2 ^a · N + 2 ^a −b becomes 0. Further, when there is an error of −b · LSB at the a-th bit from the lower order, the frequency of occurrence of the quantization level of 2 ^a · N + 2 ^a −1 is increased by b + 1 times,
Quantization levels from 2 ^a · N + 2 ^a−1 −1 to 2 ^a · N + 2
The appearance frequency of each code between ^{a-1 and} b-1 is 0.

Therefore, if the error is corrected so as to be canceled, the distortion due to the linearity error of the A / D converter can be reduced. However, the sample data, which should be a code with a low frequency of appearance, is erroneously converted into a code with a high frequency of occurrence. Therefore, among a plurality of samples having a quantization level with a high appearance frequency, a sample that should originally be another quantization level is included.

In order to correct it, it is first necessary to check at what rate the erroneous conversion is performed. It is theoretically possible to obtain a plurality of codes having an extremely high appearance frequency under the influence of one error bit detected by the error bit detection means 5. Let X be the average of the appearance frequencies at which a plurality of such codes are actually detected. Then, for a sample having a code whose occurrence frequency is extremely increased under the influence of the error bit, the code is changed to 1 in which the occurrence frequency theoretically becomes extremely small due to the one error bit.
Or, in a plurality of codes, depending on the value of X, for example,
It suffices to convert every X to one. When there are errors in a plurality of bits, the above operation is performed for each.

The amplitude level quantized under the influence of the error bit detected by the error bit detecting means 5 is
In S23 shown in FIG. 4, it is determined whether the correction should be positive or negative depending on the difference between the quantization levels of the adjacent samples on the time axis.

FIG. 5 is an A / D conversion characteristic diagram for explaining a method of estimating a correct code. In the figure, the horizontal axis represents the analog input and the vertical axis represents the digital output after A / D conversion. The broken line represents the ideal characteristic of the A / D converter. For a code having a value of 1 at the digit position of the error bit, the internal file is read to check how many samples the code is continuous on the time axis. For example, when it is determined that there is an error of b = + 1 · LSB at the a = 3rd bit from the lower order, the input / output characteristics of the A / D converter are as shown in FIG. That is, 8
At N + 3, the digital output does not change with an increase in analog input of + 1 · LSB. Therefore, 8
At N + 3, the frequency of appearance is extremely high. Then, at 8N + 6, the digital output does not become 8N + 7 and increases by 8 when the analog input increases by + 1 · LSB.
It becomes N + 8. In the characteristic diagram shown in FIG. 5, 8N + 3 to 8N is set at the output position when N = -2.
N + 6 is described.

FIG. 6 is an explanatory view showing a digital output code on the time axis for explaining a method of estimating a correct code. In the figure, the horizontal axis represents time, and the vertical axis represents the digital output code of the sample at each sampling time as an amplitude level. In order to reduce the linearity error, all the digital outputs corresponding to 8N + 4 to 8N + 6 are converted into a large value by + 1 · LSB. Then, in FIG. 5, when the samples of the digital output corresponding to 8N + 3 are K ₁ , K ₂ , ..., K _t ,
If two or more samples corresponding to 8N + 3 continue for t samples, K ₁ and the sample K ₀ one sample before that
The difference between the digital outputs of and is taken, and this difference is designated as D. D
Is positive, the value from K _{t / 2} to K _t is incremented by 1 and D
When is negative, the value of K ₁ to K _{t / 2} is incremented by +1.

By doing so, the digital output samples corresponding to 8N + 3 at a rate of once every two times are converted into a large value by + 1 · LSB. Which sample is converted is determined by D, which is the difference in quantization level between adjacent samples on the time axis. In the above description, the case where it is determined that there is an error of b = + 1 · LSB has been described, but a correct code can be similarly estimated even if the value of b is another value. If the value of t is an odd number, the value of t / 2 is rounded down or rounded up. By the above operation, A
Data converted into an incorrect value due to the linearity error of the / D converter can be easily corrected, and the linearity error can be reduced.

As a method of correcting the linearity error, various interpolation techniques used when estimating the quantization level of a sample from the quantization levels of adjacent samples on the time axis can be used. That is, consecutive samples having a code in which the error bit determined by the error bit detection means 5 influences the quantization level can be corrected by using these interpolation techniques.

In the above description, the embodiment as the linearity error reducing device of the A / D converter has been described.
Normalizing means 4 used as an example of the configuration of a part of the normalizing means 4.
The blocks up to can be used as a device for detecting the linearity error of the A / D converter.

[0035]

As is apparent from the above description, according to the present invention, even if the characteristics of the A / D converter used are not known, the music software having distortion due to the linearity error of the A / D converter is used. This has the effect of reducing distortion and improving the quality of music software. Moreover, even if the A / D converter itself is not used, this A
There is an effect that the linearity error of the / D converter can be detected.

[Brief description of drawings]

FIG. 1 is a block diagram illustrating an outline of an embodiment of the present invention.

FIG. 2 is an example of a histogram of an actual digital audio signal.

FIG. 3 is a flowchart illustrating processing steps for realizing the function of error bit detection means.

FIG. 4 is a flowchart showing a procedure for correcting a linearity error on a time axis.

FIG. 5 is an A / D conversion characteristic diagram for explaining a method of estimating a correct code.

FIG. 6 is an explanatory diagram showing a digital output code on a time axis for explaining a method of estimating a correct code.

[Explanation of symbols]

1 ... Switching means, 2 ... Memory, 3 ... Histogram forming means, 4 ... Normalizing means, 5 ... Error bit detecting means, 6 ... Digital signal processing means.

Claims (2)

[Claims] 1. An appearance frequency detecting means for inputting at least a part of a digital audio signal converted into a code by an A / D converter and detecting an appearance frequency of the code, and inputting an output of the appearance frequency detecting means, the A Linearity error extracting means for extracting linearity error of A / D converter, error reduction for reducing distortion due to linearity error of the A / D converter based on output of the digital audio signal by inputting the digital audio signal A linearity error reducing device for an A / D converter, characterized in that it has means. 2. An appearance frequency detecting means for inputting at least a part of a digital audio signal converted into a code by an A / D converter and detecting an appearance frequency of the code, and inputting an output of the appearance frequency detecting means, A linearity error for extracting the linearity error of the A / D converter as a characteristic of the appearance frequency by dividing the appearance frequency of each code by the average appearance frequency of each code and the code whose quantization level is in the vicinity of each code. A linearity error detection device for an A / D converter, characterized in that it has extraction means.