JP3219169B2 - Digital audio signal processor - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital audio signal processor for decoding a PCM code into an analog audio signal.

[0002]

2. Description of the Related Art In a digital audio transmission system in which an analog audio signal is encoded into a PCM code on a transmission side and transmitted to a transmission line, and the reception side decodes the PCM code into an analog audio signal, an error occurs in the PCM code on the transmission line. Is added, the output (reproduced) audio waveform on the receiving side is different from the input audio waveform on the transmitting side, so that an abnormal sound may be generated and the quality of the audio is deteriorated. there were.

Conventionally, as a device for solving such a problem, received PCM codes are always accumulated for a certain time length, and when an error is detected, the PCM code of that portion is immediately stored. There is a device that replaces an error-free PCM code.

FIG. 2 shows an example of this type of conventional apparatus. In the figure, reference numeral 1 denotes a buffer memory for PCM code processing,
Is a smoothing filter, 3 is an output buffer controller,
Reference numeral 4 denotes a silent signal generator.

In the above configuration, the received PCM code (sound signal) a is always stored in the PCM code processing buffer memory 1 by an amount corresponding to a fixed time length. On the other hand, an error in the PCM audio signal a is checked by an error detector (not shown), here a CRC error detector. If no error is detected, the CRC error detection signal b is off and the output buffer control The unit 3 turns off the control signals c and d to the buffer memory 1 for PCM code processing and the smoothing filter 2 so as not to perform the process for removing abnormal noise. Therefore, in this case, the output PCM audio signal e is a signal that is delayed from the input PCM audio signal a by the delay time in the PCM code processing buffer memory 1 and the smoothing filter 2.

On the other hand, if an error is detected in the PCM audio signal a, the CRC error detection signal b is turned on, and the output buffer control unit 3 switches control signals c and d to perform processing for removing abnormal noise. Is performed. Specifically, the PCM code processing buffer memory 1 stores the immediately preceding PCM
Since the audio signal is stored, the PCM audio signal considered to be abnormal due to an error is
Replace with CM audio signal. Further, by performing such a replacement process, a discontinuous portion of the PCM audio signal occurs before and after the processed portion. Therefore, the smoothing filter 2 is applied only to the discontinuous portion to smooth the waveform.

When an error of a long period (about 20 msec or more) that cannot be compensated for in the above-described PCM code replacement process is detected continuously, a silent signal f is generated by the silent signal generator 4. Thus, the output signal g of the buffer memory 1 for PCM code processing is replaced.

The PCM audio signal e that has been subjected to the above processing is
By sending the signal to the / A converter and decoding it into an analog audio signal, generation of abnormal noise is prevented, and the audio quality is improved.

[0009]

By the way, if the code error is about one bit of the PCM code composed of a plurality of bits, the reproduced waveform is hardly affected even if decoded as it is, and no abnormal noise is generated. There are cases.

In the above-described conventional digital audio signal processing apparatus, since the above-described replacement processing is performed based on the CRC error detection signal, the processing is performed even for a one-bit error. However, there is a problem that the voice quality may be greatly deteriorated.

SUMMARY OF THE INVENTION In view of the above-mentioned conventional problems, the present invention is to improve the quality of speech without fail by processing only those of the code errors added on the transmission path which affect the speech signal at the time of decoding. It is an object of the present invention to provide a digital audio signal processing device capable of performing the following.

[0012]

In order to achieve the above object, according to the present invention, in a digital audio signal processing apparatus for decoding a PCM code into an analog audio signal, an absolute value of a difference between adjacent PCM codes is provided. Means, and means for comparing the absolute value of the difference and a predetermined threshold,
A digital audio signal processing device comprising means for performing abnormal noise removal processing only when an error is detected in a PCM code and the absolute value of the difference exceeds a predetermined threshold.

According to a second aspect of the present invention, there is provided a digital audio signal processing device for decoding a PCM code into an analog audio signal, means for detecting whether the PCM code continuously exceeds a predetermined value, And a means for performing abnormal noise removal processing only when an error is detected and the PCM code continuously exceeds a predetermined value.

According to a third aspect of the present invention, there is provided a digital audio signal processing apparatus for decoding a PCM code into an analog audio signal, means for obtaining an absolute value of a difference between adjacent PCM codes, Means for comparing whether the PCM code continuously exceeds a certain value, and means for detecting whether an error is detected in the PCM code and the absolute value of the difference exceeds a predetermined threshold. Or the PC
A digital audio signal processing device comprising means for performing abnormal noise removal processing only when the M code continuously exceeds a certain value.

[0015]

According to the first aspect of the present invention, the received PCM
For the code, the absolute value of the difference between adjacent PCM codes is obtained, the absolute value of the difference is compared with a predetermined threshold, an error is detected in the PCM code, and the absolute value of the difference exceeds a predetermined threshold. At this time, an abnormal noise removal process is performed.

Further, according to the second aspect, the received PC
It is detected whether or not the M code continuously exceeds a certain value. If an error is detected in the PCM code and the PCM code continuously exceeds a certain value, abnormal noise removal processing is performed. Will be

Further, according to the third aspect, the received PC
For the M code, the absolute value of the difference between adjacent PCM codes is obtained, the absolute value of the difference is compared with a predetermined threshold value, and whether the received PCM code continuously exceeds a certain value is determined. Is detected, when an error is detected in the PCM code and the absolute value of the difference exceeds a predetermined threshold, or when the PCM
When the code continuously exceeds a certain value, abnormal noise removal processing is performed.

[0018]

FIG. 1 shows an embodiment of a digital audio signal processing apparatus according to the present invention. In FIG. 1, the same components as those of the prior art are denoted by the same reference numerals. That is, 1 is a buffer memory for PCM code processing, 2 is a filter for smoothing, 3 is an output buffer controller, 4 is a silence signal generator, 5 is a register for delay, 6 is a subtractor, 7 is an absolute value comparator, 8 Denotes an overload detection unit, 9 denotes an OR gate, and 10 denotes an AND gate.

In the above configuration, the received PCM audio signal a is always stored in the PCM code processing buffer memory 1 by an amount corresponding to a fixed time length,
The subtracter 6 calculates the difference between the PCM audio signal h delayed by the delay register 5 for the sample. This PCM
The difference value i of the audio signal is converted to an absolute value, and a predetermined threshold value j
Is compared with the absolute value comparing unit 7. When the difference value i exceeds the threshold value j, the difference comparison signal k turns on.

On the other hand, the received PCM audio signal a is also sent to the overload detecting section 8, and it is determined whether or not a constant value, here the maximum value that can be represented by the PCM code, continues for two or more samples. When the detection is performed by the unit 8 and two or more samples are continuous (overload state), the overload detection signal 1 is turned on.

The difference comparison signal k and the overload detection signal l
Is ORed by an OR gate 9 to generate an abnormal noise detection signal m, and further ANDed by an AND gate 10 with a CRC error detection signal b to generate an abnormal noise processing signal n.
It is input as a control signal of the output buffer control unit 3. The output buffer controller 3 controls the on / off of the control signals c and d in accordance with the on / off of the abnormal sound processing signal n, as in the case of the conventional example.

In order to show the effectiveness of using the difference value of the PCM code for abnormal noise detection, a 14-bit (bit) linear P
FIG. 3 shows the relationship between the threshold value when the CM code is used and the probability of occurrence of samples exceeding the threshold value. Here, using four voice samples of males and females, statistical results of a case where there is no error in the ADPCM code (error free) and a case where an error is detected in the ADPCM code are shown. Since it is considered that the sound of the CRC error burst contains more abnormal sounds than the normal sound, the difference in the occurrence probability in FIG. 3 can be applied to abnormal sound detection.

According to FIG. 3, the sound containing no abnormal sound contains about 90% within the difference value 200, while the sound containing the abnormal sound contains only about 60% within the difference value 200. You can see that. If no abnormal noise occurs even if a CRC error is detected, the probability of occurrence of the difference value in the burst is considered to be the same as in the case of error free, and the probability of erroneous detection includes the case where abnormal noise is included. The number of unnecessary processes can be reduced.

The specific noise removal processing may be the same as that of the conventional example as described above, but the burst immediately preceding the burst in which the abnormal noise is detected is also subjected to the abnormal noise removal processing. This eliminates the need to detect the rising portion of the abnormal sound, so that it is possible to increase the threshold value of the difference value of the PCM code. As described above, the false detection of the normal sound is reduced, and the abnormal sound processing is performed. Efficiency is improved.

Since the feature of the present invention resides in the detection (selection) of the noise removal part, a method other than those described here may be used as a processing method for noise removal.

FIG. 4 shows another embodiment of the present invention, in which an example using another abnormal noise elimination method is shown.
An A converter, 12 is an attenuator, and 13 is a switch. In the above configuration, the received PCM audio signal a is converted into an analog audio signal o by the D / A converter 11, input to the switch 13 as it is, and attenuated by the attenuator 12 (p) and input. At this time, if the abnormal sound is not detected and the abnormal sound processing signal n is in the off state, the switch 13 selects and outputs the unattenuated analog audio signal o. If the sound processing signal n is on, the switch 13 selects and outputs the attenuated analog audio signal p. As described above, by suppressing the output level when the abnormal sound is generated, the sound quality can be improved.

In the above description, the difference comparison signal k
And an abnormal noise detection signal m by taking a logical sum of the
Although a may be abnormal sound detection signal m by using only the overload detection signal l.

[0028]

As described above, according to the present invention, a signal indicating whether or not a difference value between adjacent PCM codes exceeds a predetermined threshold value or whether or not a PCM code is overloaded is determined. The abnormal noise processing signal is obtained by the logical product of the logical sum signal of these signals or both of them and the error detection signal, and the abnormal noise removal processing is controlled. In addition, there is less processing of bursts having errors but having little effect on voice quality, that is, bursts whose voice quality is degraded by performing abnormal noise removal processing. As a result, voice quality is improved and unnecessary The number of processes is reduced.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing one embodiment of a digital audio signal processing device of the present invention.

FIG. 2 is a configuration diagram showing an example of a conventional digital audio signal processing device.

FIG. 3 is a graph showing a relationship between a threshold value for a difference value of a PCM code and an occurrence probability of a sample exceeding the threshold value;

FIG. 4 is a block diagram showing another embodiment of the digital audio signal processing device of the present invention.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Buffer memory for PCM code processing, 2 ... Filter for smoothing, 3 ... Output buffer control part, 4 ... Silence signal generation part, 5 ... Register for delay, 6 ... Subtractor, 7 ... Absolute value comparison part, 8 ... Over Load detector, 9: OR gate, 10: AND gate.

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Shuzo Kato 1-6-6 Uchisaiwaicho, Chiyoda-ku, Tokyo Nippon Telegraph and Telephone Corporation (56) References JP-A-61-108233 (JP, A) JP-A Sho 61-163743 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) H04B 14/00-14/06 H04L 1/00

Claims (2)

(57) [Claims] 1. A digital audio signal processing device for decoding a PCM code into an analog audio signal, wherein the maximum value that the PCM code can represent with the PCM code is 2 samples.
Detects whether or not the sample is continuously taken for more than pull, and 2 samples or more
Means for setting an overload state when the data is continuously taken; and detecting an error in the PCM code and
Digital audio signal processing apparatus characterized by comprising a means for performing the abnormal noise removal processing only when that. 2. A digital audio signal processing device for decoding a PCM code into an analog audio signal, comprising: means for obtaining an absolute value of a difference between adjacent PCM codes; and means for comparing the absolute value of the difference with a predetermined threshold value And the maximum value that the PCM code can represent with the PCM code is 2 samples.
Detects whether or not the sample is continuously taken for more than pull, and 2 samples or more
Means for setting an overload state in the case of continuous taking; and abnormal noise removal processing only when an error is detected in the PCM code and the absolute value of the difference exceeds a predetermined threshold value or when the overload state is established. A digital audio signal processing device.