JP2003140694A - Audio decoder - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problem that it has been difficult for a user to recognize audio errors when decoding audio data containing audio errors. SOLUTION: The audio error is made recognizable for a user by adding noise, music, or speech as a predetermined signal in proportion to an error incidence rate.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an audio decoder which reproduces an encoded audio signal.

[0002]

2. Description of the Related Art Conventionally, as an audio decoder, an audio decoder capable of suppressing an abnormal sound at the time of audio reproduction due to an error occurring in a transmission line has been disclosed in Japanese Patent Laid-Open No. 8-116314.

FIG. 7 shows a block diagram of a conventional audio decoder. Reference numeral 101 denotes a storage unit for storing the data of the MPEG audio stream on a frame-by-frame basis.
Reference numeral 02 denotes a CRC check of the MPEG audio stream to detect the presence or absence of an error. When an error is detected, the frame in which the error is detected is prohibited from being written to the storage means 101, and instead the MPEG audio of the previous frame is replaced. Error processing means 103 for controlling the writing of the stream and for muting the audio signal when an error is continuously detected is the storage means 1.
Read MPEG audio stream from 01 and MPE
104 is a decoding unit that performs G audio decoding.
Is a mute means for muting the output of the decoding means 103 when the error processing means 102 controls mute.

The operation of the conventional audio decoder shown in FIG. 7 will be described below. The MPEG audio stream is CRC checked by the error processing means 102, and if no error is detected, the storage means 101
Are sequentially stored in the decoding unit 103 and sequentially decoded by the decoding means 103.
It is reproduced as a CM output signal.

On the other hand, if the error processing means has an error and the error is detected by the error processing means 102, the MPEG audio stream in which the error is detected is not stored in the storage means 101, but instead the error is detected at the end. The unrecorded MPEG audio frame is stored in the storage means. Since the MPEG audio frame read by the decoding means 103 is the correct previous MPEG audio frame, the PCM signal is repeated in the MPEG audio frame cycle, and no loud noise is generated.

Further, when an error is continuously detected, the error processing means 102 mutes the output of the decoding means 103 by the mute means 104 and outputs it as a final audio decoder output. When a certain audio frame is input to the audio decoder, no sound is output due to the silent output.

[0007]

However, in the above-mentioned configuration, when the signal goes out of the receiving area from the receiving area like the terrestrial broadcasting, the reproduction signal is muted and the silent state continues, and the user receives the signal. It is difficult to recognize that you have gone out of the area.

In view of the above problems, the present invention suppresses abnormal noise when an error occurs in audio data, and
It is an object of the present invention to provide an audio decoder capable of notifying a user that he / she has gone out of the reception area by adding a predetermined signal such as music, announcement or noise.

[0009]

In order to achieve this object, an audio decoder according to the present invention has an error detecting means for detecting an error in encoded audio data and an error detecting means for detecting an error in the audio data. When not present, the audio data is decoded and a PCM signal is output, and when an error is detected by the error detection means, a decoding means for muting and outputting the PCM signal corresponding to the audio data including the error, and a predetermined signal are generated. When the error is not detected by the signal generating means and the error detecting means, the output of the decoding means is output,
When an error is detected, it has a switch means for outputting the output of the signal generating means, and a signal level control means for controlling the output signal level of the signal generating means according to the error detection result of the error detecting means.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION (Embodiment 1) Hereinafter, Embodiment 1 of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram of an audio decoder according to the first embodiment of the present invention. In FIG. 1, reference numeral 1 is an error detecting means for detecting an error in the input encoded audio data, and 2 is a decoder means for decoding the encoded audio data. When the error detecting means 1 detects no error, the audio data is outputted. Is decoded and mute output is performed when an error is detected. Reference numeral 3 denotes a signal generating means for generating a predetermined signal, and 4 denotes an output of the signal generating means 3 when the error detecting means 1 detects an error, and selects an output of the decoder means 2 when no error is detected. The switch means 5 for controlling is a signal level control means for controlling the signal level generated by the signal generating means 3 based on the error detection result of the error detecting means 1. The signal generating means 3 generates noise such as white noise or music or an announcement.

FIG. 2 is a block diagram showing the details of the signal level control means 5, and 51 is an error occurrence rate calculation means for calculating the error detection result short-time average value of the error detection means 1.
Reference numeral 52 is a level control means, which calculates a control value of the signal level from the error occurrence rate calculated by the error occurrence rate calculation means 51 and controls the output level of the signal generation means 3.

In the embodiment, the compressed audio data is MPEG audio (ISO / IEC 11172-3, ISO / IEC13813-3, ISO / IEC13813-) having a frame structure.
7) are widely put into practical use. Error detection for compressed audio having such a frame structure includes missing frame sync words, abnormal parameters, and CRC.
It is conventionally known that the check word is used. The error detecting means 1 detects an error in audio data having a frame structure by such a conventional error detecting method.

When there is an error in the audio data and the error is not detected by the error detecting means 1, the audio frame is decoded by the decoding means 2 and output PCM.
A signal is selected by the switch means 4, and a PCM signal is output as an audio decoder output.

If the audio data has an error, the error detecting means 1 detects the error. When an error is detected, the switch means 4 selects the output of the noise generating means during the PCM output data period corresponding to the audio frame containing the error, and the noise generated by the noise generating means is output from the audio decoder. In the signal level control means 5, the error occurrence rate for a short time is calculated in the error occurrence rate calculation means 51 by using a short-time integration calculation, and the signal level is set so as to be inversely proportional to the error occurrence rate calculated by the level control means 52. Control like. The signal level output from the signal generating means 3 by controlling the signal level as shown in FIG. 3 is very small immediately after going out of the receiving area because the error occurrence rate is small,
After that, as the error occurrence rate increases, it gradually increases. In the example of FIG. 3, the maximum noise level is 30% of the maximum amplitude.
The noise level is about the same. FIG. 3 is an example of the relationship between the error occurrence rate and the signal level, and is not limited to FIG.

The error occurrence rate calculation means 51 uses short-time integration, which is easy to implement, as a method of calculating the error occurrence rate. An example is shown in the formula. Error occurrence rate Pe at time j
(J) is calculated from the error: Error (j) and the error occurrence rate Pe (j-1) at time j-1 by the following formula.

[0016]

[Equation 1]

Here, j represents the time based on the calculation cycle T0. Error (j) is the error detection result at time j, which is 1 when an error is detected and 0 when no error is detected. Tup is a rising time constant, and Tdn is a falling time constant. At least Tup and Tdn have a larger time constant than T0.

From the above equation, the fade-in speed of the signal generated by the signal generating means 3 can be controlled by the rising time constant: Tup. Similarly, the rate of fading out of the signal generated by the signal generating means 3 can be controlled by the falling time constant: Tdn. For example, in the case of an audio decoder for reproducing music in terrestrial broadcasting, these time constants Tup and Tdn are used as signals generated by the signal generator 3 at an appropriate time, for example, noise or a predetermined music signal. By setting Tup to 3 seconds and Tdn to 1 second, when the user goes out of the receiving area, the user recognizes that he / she went out of the receiving area within a few seconds by noise or increase of a predetermined music signal. In addition, when entering the receiving area from outside the receiving area, it is possible to know that the inside of the receiving area is entered by noise or a predetermined music signal level reduction in about 1 second.

In the first embodiment, the audio decoder outputs audio data as an error continues and an error occurrence rate increases without generating an offensive noise to the audio data in which an error is detected. To PC
The user can be notified by noise or a predetermined music signal that the M signal cannot be reproduced. Since the signal level increases corresponding to the error occurrence rate, the error occurrence rate is small for sudden errors, so very small noise or music signals are output from the signal generating means. It does not disturb the person.

(Second Embodiment) FIG. 4 is a block diagram of an audio decoder according to the second embodiment of the present invention. In FIG. 4, 1 is an error detecting means for detecting an error in the input encoded audio data, 2 is a decoder means for decoding the encoded audio data, and when the error detecting means 1 does not detect an error, the audio data Is decoded and mute output is performed when an error is detected. Reference numeral 3 is a signal generating means for generating noise, and the maximum signal level is set smaller than the maximum output level of the decoding means 2. For example, the maximum signal level is about 30% of the maximum output level of the decoding means 2. Reference numeral 6 is a mixer means for adding the respective output signals of the decoding means 2 and the signal generating means 3 while weighting them, and 7 is for controlling the weighting gain of the mixer means 6 by using the error occurrence rate detected by the error detecting means 1. It is a gain control means.

FIG. 5 is a block diagram showing the details of the gain control means. 61 is an error occurrence rate calculation means for calculating the occurrence rate of the error detected by the error detection means 1 for a certain period, 6
2 is a first gain calculating means for calculating a gain G1 to be added to the output signal of the decoding means 2, and 63 is a signal generating means 3.
Is a second gain calculating means for calculating a gain G2 to be added to the output signal of.

In the second embodiment, the output PCM signal of the decoding means 2 and the output signal of the signal generation means 3 are added by the mixer means 6. The first gain G1 with respect to the output PCM signal of the decoder means 2 and the second gain G2 with respect to the output signal of the signal generation means 3 are the error occurrence rate Pe calculated by the error occurrence rate calculation means 61 and as shown in FIG. Get involved. When the error occurrence rate Pe is 0, the gain G2 is 0 and the signal generated by the signal generating means 3 is the mixer means 6
Is not added, and only the output PCM signal of the decoding means 2 is output from the mixer means 6. Error rate is 1
As the first gain G1 decreases, the first gain G1 decreases and the second gain G2 increases. Thus, as the error occurrence rate increases, the signal generated by the signal generating means 3 is output from the output of the mixer means 6 at a higher level, and
The PCM output signal of the decoder means 2 gradually decreases,
It will be muted.

The second embodiment differs from the first embodiment in that the PCM output of the decoder means 2 fades out as the error occurrence rate Pe increases. As the error occurrence rate increases, the noise and the music signal level generated by the signal generating means gradually increase, and the PCM signal decoded by the decoding means 2 is reproduced at a large level in a sporadic manner, resulting in a loud and unpleasant sound. Can be prevented.

The error occurrence rate calculation means 61 uses the short time integration of the same formula as in the first embodiment. The rising time constant Tup can control the fade-out speed of the PCM output of the decoding means 2, the noise generated by the signal generating means 3, and the fade-in speed of the music signal. Similarly, the rate of fading in of the PCM output of the decoding means 2 and fading out of noise and music signals generated by the signal generating means 3 can be controlled by the fall time constant Tdn.
It is possible to control the switching speed of the output PCM signal and the noise generated by the signal generating means 3 and the music signal while cross-fading.

In the case of an audio decoder for reproducing music by terrestrial broadcasting, these time constants Tup and Td are set.
By setting n to an appropriate time, for example, Tup is set to 3 seconds and Tdn is set to 1 second, noise or a music signal generated from the reproduced sound signal generating means 3 in about 3 seconds when going out of the reception area. When the level increases and the user recognizes that he / she has gone out of the reception area, and when he / she goes into the reception area from the outside of the reception area, the noise or the music signal generated from the signal generation means 3 in about 1 second. It is possible to know that the level has decreased, the PCM output signal correctly decoded by the decoding means has increased, and the signal has entered the reception area.

As the signal generated by the signal generating means, white noise having a flat spectrum or noise having the same spectrum as the statistical spectrum of music is used to reproduce the sound reproduced outside the reception area in the conventional analog broadcasting. The same sound can be used, and it becomes easier for the user to recognize that the user is outside the reception area. Similarly, even with predetermined music, the user can more easily recognize that the music is outside the reception area, and it is less annoying than noise. In addition, it is possible to clearly notify the user that the audio cannot be reproduced outside the reception area or due to the weak radio wave by using the voice announcement.

[0027]

As described above, according to the first aspect, the error detecting means for detecting an error in the encoded audio data and the audio data when the error detecting means detects no error in the audio data. Decoding means for decoding and outputting a PCM signal, muting and outputting a PCM signal corresponding to audio data containing an error when an error is detected by the error detecting means, signal generating means for generating a predetermined signal, and error The switch means for outputting the output of the decoding means when the error is not detected by the detecting means, and the output of the noise generating means when the error is detected, and the output signal level of the signal generating means for detecting the error of the error detecting means An error occurs due to the signal level control means for controlling according to the result, and the audio decoder That it can no longer reproduce the PCM signal from the audio data noise or predetermined music, there is the advantage that it is possible to inform the user by announcements. In addition, the level of the signal generated by the signal generating means increases in accordance with the error occurrence rate, so that very small noise, music, or voice is generated against an unexpected error, which is offensive to the user. It never happens.

According to the second aspect, the calculation of the error occurrence rate is performed by using the short-time integration, and the rising time constant and the falling time constant of the integration are individually set, so that the noise level increasing speed and the noise speed decreasing speed are separately set. Can be controlled.

According to a third aspect of the present invention, an error detecting means for detecting an error in the encoded audio data, and when no error is detected in the audio data by the error detecting means, the audio data is decoded and a PCM signal is output. When an error is detected by the error detecting means, a decoding means for muting and outputting a PCM signal corresponding to audio data including the error, a signal generating means for generating a predetermined signal, an output of the decoding means and a signal generating means An error occurs and the audio decoder cannot reproduce the PCM signal from the audio data due to the mixer means for adding the respective outputs of the respective weights while adding them and the gain control means for controlling the weighting amount of the mixer means according to the error detection result of the error detection means. That things like noise, music, voice, etc. It is possible to inform the user by a predetermined signal, a high status error occurrence rate is reproduced as a large amplitude of a PCM signal when decoded by the decoding means single correct audio decoder,
There is an advantage that it is possible to prevent a loud noise that is offensive to the ear.

According to the fourth aspect, the calculation of the error occurrence rate is performed by using the short-time integration, and the rising time constant and the falling time constant of the integration are individually set, so that the PCM signal decoded by the decoding means is obtained. It is possible to control the switching speed while crossfading with the level generated by the signal generating means.

According to the fifth aspect of the present invention, by setting the rising time constant of the short-time integration in the error occurrence rate calculating means to be larger than the falling time constant, the reception state becomes poor and the signal generated by the signal generating means fades. When the reception condition is improved and the error occurrence rate is reduced, the speed at which the signal generated by the signal generating means fades out can be increased and the signal generated by the signal generating means can be quickly attenuated. Also, the decoding means PC
The fade-in speed of the M output becomes faster, the PCM signal can be quickly reproduced, the reproduced sound can be auditioned more quickly, and the signal generated by the signal generating means has a large level for a sporadic error. It doesn't get harsh and doesn't hurt.

According to the sixth aspect, by making the spectrum of the predetermined signal output by the signal generating means similar to the statistical spectrum of the music signal, the sound reproduced outside the reception area in the conventional analog broadcasting is obtained. The same sound can be used, and it becomes easier for the user to recognize that the user is outside the reception area.

According to the seventh aspect, by using the predetermined signal output from the signal generating means as music, it is easier for the user to recognize that the user is outside the reception area, as in the case of noise. It's not an offensive sensation.

According to the eighth aspect, by using a predetermined signal output from the signal generating means as a voice announcement, it is clearly notified to the user that the audio cannot be reproduced outside the receiving area or the radio wave is weakened. You can also do it.

[Brief description of drawings]

FIG. 1 is a block diagram of an audio decoder according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing details of signal level control means in the first embodiment of the present invention.

FIG. 3 is a diagram showing an example of a relationship between an error occurrence rate and a signal level generated by a signal generating means in Embodiment 1 of the present invention.

FIG. 4 is a block diagram of an audio decoder according to the second embodiment of the present invention.

FIG. 5 is a block diagram showing details of gain control means in the second embodiment of the present invention.

FIG. 6 is a diagram showing an example of a relationship between an error occurrence rate and a gain according to the second embodiment of the present invention.

FIG. 7 is a block diagram of a conventional audio decoder.

[Explanation of symbols]

1 Error detection means 2 Decoding means 3 Signal generation means 4 switch means 5 Signal level control means 6 Mixer means 7 Gain control means 51 Error occurrence rate calculation means 52 Level control means 61 Error occurrence rate calculation means 62 First Gain Calculation Means 63 Second gain calculating means 101 storage means 102 error handling means 103 decoding means 104 Mute means

Claims (9)

[Claims] 1. Decoding encoded audio data to obtain P
An audio decoder for outputting a CM signal, comprising: error detecting means for detecting an error in the encoded audio data; and decoding the audio data when no error is detected in the audio data by the error detecting means. Decoding means for outputting a PCM signal, muting and outputting a PCM signal corresponding to audio data including the error when the error detecting means detects an error, signal generating means for generating a predetermined signal, and Switch means for outputting the output of the decoding means when no error is detected by the error detecting means, and for outputting the output of the signal generating means when an error is detected, and a signal level of the signal generating means for detecting the error The error rate of the instrument was calculated for a certain period, and the calculated error An audio decoder comprising a signal level control means for controlling a signal level so as to be inversely proportional to the error rate. 2. The signal level control means comprises an error occurrence rate calculation means and a level calculation means for calculating a signal level inversely proportional to the error occurrence rate calculated by the error occurrence rate calculation means, The error occurrence rate calculation means calculates the error occurrence rate for a certain period by short-time integration, and the time constant for short-time integration differs depending on whether the error occurrence rate increases or decreases. The audio decoder according to claim 1, wherein: 3. Decoding encoded audio data to obtain P
An audio decoder for outputting a CM signal, comprising: error detecting means for detecting an error in the encoded audio data; and decoding the audio data when no error is detected in the audio data by the error detecting means. Decoding means for outputting a PCM signal, muting and outputting a PCM signal corresponding to audio data containing the error when an error is detected by the error detecting means, signal generating means for generating a predetermined signal, and Mixer means for adding gains to the output of the decoding means and the output of the signal generating means, and a weighting amount of the mixer means for calculating the error rate of the error detecting means for a certain period, and calculating Output of the decoding means in inverse proportion to the error occurrence rate Audio decoder, characterized in that to control the gain, which is composed of the gain control means for controlling a gain for the output of said signal generating means in proportion to the calculated error rate against. 4. The gain control means includes an error occurrence rate calculation means, a first gain calculation means for controlling a gain with respect to an output of the decoding means, and a second gain calculation for controlling a gain with respect to an output of the signal generation means. The error occurrence rate calculation means calculates the error occurrence rate for a certain period by short-time integration, and the time constant for short-time integration is such that the error occurrence rate increases and the error occurrence rate increases. 4. The difference in the case of decrease.
The described audio decoder. 5. The error occurrence rate calculation means calculates the error occurrence rate by short-time integration, and the time constant for short-time integration is the time constant when the occurrence rate increases. 5. The audio decoder according to claim 2, wherein the time constant is larger than the time constant when it decreases. 6. The audio decoder according to claim 1, wherein said signal generating means is a noise signal having a statistical spectrum of a music signal. 7. The audio decoder according to claim 1 or 2, wherein the signal generating means is a music signal. 8. The audio decoder according to claim 1, wherein said signal generating means is a voice announcement. 9. The audio decoder according to claim 1, wherein the audio data is audio data having a frame structure.