JP2002014698A - Audio reproducing device, and video and audio decoding device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make easily restorable data by interpolating data that are computed while considering the flow of before and after data signals. SOLUTION: The device is provided with a separating means which separates headers, bit distribution, error detecting words, scale factor selection information, scale factors and samples from a bit stream, an error detecting means which conducts error detection of the separated data, a scale factor decoding means which decodes the separated scale factors, a bit distribution decoding means which decodes the separated bit distribution, an inverse quantizing means which conducts inverse quantization by employing the separated samples and the bit distribution decoded by the bit distribution decoding means, a subband combining means which combines subband data computed by the inverse quantizing means and outputs PCM data, a buffer means which stores the PCM data and a control means which adaptively restores the PCM data of the buffer means by the signals from the error detecting means and outputs the restored data.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound reproducing apparatus having an error detecting function, and more particularly, to a DVD camera, an MPEG (MPEG),
The present invention relates to an audio reproducing apparatus suitable for application to an image and audio reproducing apparatus such as a camera, an electronic still camera, and a mobile communication terminal.

[0002]

2. Description of the Related Art By digitally recording an audio signal, the reproduced sound quality is no longer affected by deterioration characteristics inherent to the recording medium, but is more susceptible to data errors due to dust and scratches on the medium. As the ASCII "DVD and Video CD" Hiroshi Fujiwara found, to solve them, Reed-Solomon code and CRC (Cyclic Redunda
ncy Code) is often used.

In ISO / IEC 11172-3: 1993 (E) (MPEG1 audio standard; hereinafter referred to as MPEG audio standard), CRC
Error detection using is specified as an option. However, as a countermeasure for detecting an error in the audio stream, it is only recommended to repeat or mute the data of the previous frame. Also,
DVD Application Format used in DVD cameras, etc.
for DVD RewritableDiscs / Part1 VIDEO RECORDING ve
According to the r.1.0 standard, CRC is essential when using MPEG audio.

[0004] In addition, Japanese Patent Laid-Open Publication No. Hei 6-120908 discloses a decoding apparatus in which when an information error is found by error correction, the state is restored to the most probable state using probability.

[0005]

However, according to the method of dealing with error detection using a CRC defined in the MPEG audio standard, if an error is detected in several frames, the same frame is repeated during that time. Or silence continues for a while. Further, the decoding device described in Japanese Patent Laid-Open No. 6-120908 involves complicated processing using probabilities.

The inventors of the present application have previously disclosed in Japanese Patent Application Laid-Open No. H11-308
Japanese Patent No. 442 proposes a simple but effective method of calculating an interpolation point from signals before and after by making good use of an extrapolation point. However, the idea of repairing PCM data of an erroneous frame is proposed. If it is used, complicated processing is not required at the time of data restoration.

[0007] Further, an ordinary audio reproduction system often uses an output buffer, and by using this buffer, an increase in memory can be suppressed.

An object of the present invention is to make it possible to recover erroneous data without performing complicated processing in a sound reproducing apparatus having an error detecting function.

[0009]

In order to achieve the above object, an audio reproducing apparatus according to the present invention comprises a header, a bit allocation, an error detection word, scale factor selection information, a scale factor selection from a sub-band coded bit stream. Separating means for separating the samples, error detecting means for performing error detection on the data separated by the separating means, scale factor decoding means for decoding the scale factor separated by the separating means, and bits separated by the separating means. Combining bit allocation decoding means for decoding the allocation, inverse quantization means for performing inverse quantization based on the separated samples and bit allocation decoded by the bit allocation decoding means, and sub-band data calculated by the inverse quantization means Sub-band combining means for outputting PCM data; Take a Eteoku buffer means, and control means for outputting adaptively repair PCM data buffer means by a signal from the error detection unit, a configuration equipped.

In the present invention, the data can be easily restored by interpolating the data calculated with reference to the flow of the data signal before and after. Further, by sharing the buffer used at the time of restoration and the buffer used at the time of normal reproduction, memory resources can be saved.

[0011]

Embodiments of the present invention will be described below with reference to the drawings. In the following embodiment, M
The following describes an example of data restoration when a CRC error of PEG audio is detected.

FIG. 1 is a block diagram showing a configuration of an audio reproducing apparatus according to one embodiment of the present invention. In the figure,
100 is an input unit to which an audio stream is input, 10
1 is a bit stream decomposing unit that decomposes the audio stream input from the input unit 100, 102 is an error detecting unit that performs error detection, 103 is a scale factor decoding unit that decodes the scale factor decomposed by the decomposing unit 101,
104 is a bit allocation decoding unit for converting the bit allocation decomposed by the decomposing unit 101 into the actual number of bits of the sample from a prescribed table or the like; 105 is an inverse quantization unit for dequantizing the sample decomposed by the decomposing unit 101 , 106 are subband synthesizing units for synthesizing the subband data calculated by the inverse quantization unit 105, 107 is a control unit for controlling the output based on the signal from the error detecting unit 102, and 108 is a buffer for storing PCM data And 109, an output unit connected to a D / A converter or the like.

The configuration shown in FIG. 1 corresponds to a normal MPEG audio decoding unit. Layer is II. First, an audio bit stream is input to the input unit 100. The input bit stream is subjected to a header, bit allocation, CRC
Decomposed into words, scale factor selection information, scale factors, and samples. The decomposed bit allocation is converted by the bit allocation decoding unit 104 into the actual number of bits of the sample. The decomposed scale factor is decoded by the scale factor decoding unit 103 with reference to the decomposed scale factor selection information. At the same time, the error detection unit 102 performs error detection on a part of the header, bit allocation, and scale factor selection information. At this time, error detection using CRC-16 is used, and G (X) = X ¹⁶ + X ¹⁵ + X ² +1 This generator polynomial is used.

The CRC word separated by the separating unit 101 is a CRC word generated by the generator polynomial at the time of encoding. Further, a CRC word is generated by passing a part of the header separated by the separation unit 101, the bit allocation, and the scale factor selection information through the generator polynomial. This generated CR
Whether or not there is an error is determined based on whether or not the C word and the CRC word separated by the separation unit 101 match. When an error is detected by the error detection unit 102, a signal is sent to the control unit 107.

Next, using the actual number of bits calculated by the bit allocation decoding unit 104 and the scale factor decoded by the scale factor decoding unit 103, the decomposed sample is inversely quantized by the inverse quantization unit 105. I do. The dequantized sub-band data is converted to PCM data through the sub-band synthesizing unit 106. Subband synthesis section 106
Is output from the buffer unit 108 once.
Is stored in Normally, an MPEG playback system uses a buffer to absorb processing time in order to synchronize video and audio. This buffer is often a ring buffer.

If the error is not detected by the error detecting unit 102, the control unit 107 outputs the PCM data from the buffer 108 to the output unit 109 for each frame or in units of several frames. On the other hand, if an error has been detected by the error detection unit 102, the data calculated with reference to the signal flow (with reference to the signal flow before and after the error-detected signal) is interpolated and interpolated. , PC
The M data is output to the output unit 109. Hereinafter, a method of outputting PCM data with reference to a signal flow will be described.

This will be described with reference to FIGS. FIG. 2 is a detailed diagram of the buffer unit 108. Reference numerals 200 to 204 each denote a memory area for one audio frame. Therefore, five frames are shown. PXY (X = 0 to 4, Y = 0 to 3) represents signed 16-bit PCM data.

FIG. 3 is a diagram for explaining a method of interpolating data calculated with reference to the flow of signals before and after a signal in which an error is detected. The horizontal axis in FIG. 3 represents time, and the vertical axis represents the signal level. Now, signal 301 and signal 303
It is assumed that an error is detected in the signal between and. 302
Is linear interpolation data calculated from 301 and 303 at an intermediate time point between 301 and 303. Also, 305
Is the extrapolated point of 300 and 301 at the intermediate point between 301 and 303 calculated from 300 and 301, and 306 is the extrapolated point of 301 and 303 calculated from 303 and 304.
Extrapolated point between 303 and 304 at an intermediate point between the two. Referring to the signal flow, for example, as shown in FIG. 3, when obtaining a signal for restoration between 301 and 303 (for interpolating between 301 and 303), 3
This means that 305 is the most optimal level as a signal, rather than 02 or 306.

This will be described in more detail.
Normally, the level change of the analog audio signal along the time axis indicates a smooth change such as a sine wave, so in the case of the example shown in FIG. Therefore, the level change of the signal is not natural. That is, as shown in FIG.
The rate of increase of the signal level of one (of the previous two data);
In the case where the polarity of the signal level increase ratio of the 303 and 304 (the last two data) is opposite (the polarity of the slope ta and the polarity of the slope tb are opposite), and the arrangement of the data (signal) is upwardly convex. Should be interpolated (interpolated) between 301 and 303
The signal level of the data is higher when the signal level is higher than the linear interpolation point 302 (that is, the linear interpolation point 3 in the example of FIG. 3).
02 above) is more natural in terms of signal flow. Also, comparing the two extrapolated points 305 and 306, the one with the smaller signal level (here, extrapolated point 305
Does not show extreme protrusions and has a smoother transition,
It is more natural in terms of signal flow. This is the reason that in the case of the example of FIG. 3, the signal 305 is the most optimal signal as compared with the signal 302 or 306.

According to the above idea, the information up to two frames ahead of the frame in which the error was detected is used. Therefore, data is written into the buffer unit 108 at least two frames ahead of the frame in which the error is detected. deep.

The case where an error is detected in the frame 202 will be described below. In this case, 200 and 20
1. In addition, the PCM data in the frame 202 is repaired from the PCM data in the frames 203 and 204. First, it is assumed that P20 of the frame 202 is obtained. In FIG. 3, 300 is P00 and 301 is P10.
In addition, when 303 corresponds to P30 and 304 corresponds to P40, data of 302, 305, and 306 are respectively DATA (302) and DATA (30).
5) and DATA (306), each data DATA
(302), DATA (305), DATA (306)
Is: DATA (302) = (P10 + P30) / 2 Expression (1) DATA (305) = P10 + ta * 1 Expression (2) DATA (306) = P30 + tb * 1 Expression (3) ), (2), and (3). In the equation (2), ta is a slope between 300 and 301, and is expressed as ta = (P10−P00) / 1, and tb
Is the slope of 303 and 304, tb = (P30-P40)
/ 1.

As shown in FIG. 3, when the arrangement of the signals is convex upward, DATA (305) having a lower level is selected from DATA (305) and DATA (306). By doing so, data suitable for the signal flow can be obtained. That is, when P20 is output to the output unit 109, the data (3
05) is output. Similarly, P21,
By calculating and outputting DATA at the times of P22, P23,..., It is possible to repair all the PCM data of the frame in which the error is detected. Output unit 10
The output from 9 is sent to a speaker through a D / A converter or the like.

In FIG. 3, the case where the arrangement of the signals is upwardly convex has been described. Regarding the selection of the extrapolation point at that time, DATA (305) and DAT
A (306) having the higher level may be selected.

When the arrangement of the signals is other than a convex upward or downward convex, for example, when the slopes ta and tb have the same polarity, the linear interpolation of 302 obtained from 301 and 303 is performed. The data may be repair data to be interpolated (interpolated) between 301 and 303.

The case where the repair data to be interpolated is one point has been described above. However, the present invention is not limited to one point, and it is possible to generate repair data of a plurality of points by performing the same processing at two or more points. It is.

FIG. 4 shows a case where the repair data to be interpolated is three points. Now, assume that an error is detected in three signals between the signal 401 and the signal 405. 410, 41
1, 412 are 401 calculated from 401 and 405.
And 405 are linearly interpolated data (at time points T1, T2, and T3) at the time of dividing into four equal parts. 420, 421, and 422 are extrapolated points of 400 and 401 calculated from 400 and 401, respectively, corresponding to the time points T1, T2, and T3. Also, 430,
431, 432 are the T1s from 405 and 406
Extrapolated points of 405 and 406 corresponding to the time point, T2 time point, and T3 time point, respectively. In this case, the paired extrapolation points at the time points T1, T2, and T3 are compared with each other, and the smaller signal level is determined at the time points T1, T2, T3.
Select as data to be interpolated (interpolated) at the time. That is, here, 420, 421, and 432
Is the data for restoration.

In the above description, CRC detection is used for error detection.
Is used as an example, but the present invention is not limited to CRC, and the same effect can be obtained by using other error detection.

As described above, if the data that is in the signal flow adaptively is used as the repair data, the data can be repaired relatively easily and in a natural manner along the signal flow. .

Next, another method of interpolating the data calculated with reference to the flow of signals before and after the error-detected signal in the present embodiment will be described with reference to FIG. FIG. 5 is an example of a case where an error is detected in the signal between the signal 301 and the signal 303 as in FIG.
5, the same components as those in FIG. 3 are denoted by the same reference numerals.

In FIG. 5, reference numerals 500 and 301 represent
Linear interpolation point 302 and the smaller 3
This is a linear interpolation point between 302 and 305 calculated from the extrapolation point 305. In the example shown in FIG. 3, 305 is selected as data for restoration between 301 and 303 (for interpolating between 301 and 303). Even at a signal level of 305, it takes a somewhat natural form along the signal flow,
In 305, I hate being a little too convex. Therefore, in this example, a linear interpolation point 500 of 302 and 305 is obtained and is used as repair data. By doing so, the signal flow becomes smoother than in FIG.
It is possible to perform data repair in a more natural form as viewed from the signal flow. Note that the interpolation ratio between the interpolation points 302 and 305 can be changed, and the degree of freedom of smoothness increases.

FIG. 5 shows the case where the repair data to be interpolated is one point. However, the present invention is not limited to one point, and the same processing can be performed for two or more points.

FIG. 6 shows a case where the repair data to be interpolated is three points. In FIG. 6, the same reference numerals are given to the same parts as those in FIG. In FIG. 6, 610 is a linear interpolation point calculated from the interpolation point 410 and the extrapolation point 420 at the time T1, and 611 is calculated from the interpolation point 411 and the extrapolation point 421 at the time T2. The linear interpolation point 612 is a linear interpolation point calculated from the interpolation point 412 and the extrapolation point 432 at the time T3. In the present example shown in FIG. 6, data to be interpolated (interpolated) at time T1, time T2, and time T3 is 61
0, 611, and 612 are used, respectively.

FIG. 7 is a block diagram showing the configuration of a video / audio reproducing apparatus incorporating the audio retransmitting apparatus shown in FIG. 7, reference numeral 700 denotes a recording medium such as a DVD RAM or a hard disk; 701, a separation unit for decomposing a system stream; 702, an MPEGCODEC unit for decoding a video stream; and 703, a format conversion unit for converting the format of a decoded video signal. , 704
Is a D / A converter for converting a format-converted signal from digital to analog, 705 is an output unit connected to an external output such as an LCD, and 706 is an MPEGAUDIO for decoding an audio stream and converting it to PCM data.
707 is a D for converting PCM data into an analog signal.
An / A converter 708 is an audio output unit that amplifies and outputs an audio analog signal.

[0034] The system stream recorded on the recording medium 700 is separated into a video stream and an audio stream through the separation unit 701. The separated video stream is decompressed by the MPEGCODEC unit 702 and converted into a digital video. The decompressed digital video is processed by the format conversion unit 703 as necessary.
It is converted to a TSC signal, YC or RGB signal. The format conversion unit 703 also multiplexes the OSD. The format-converted reproduction signal is converted into an analog signal by the D / A conversion unit 704, and is output to an LCD, an external output terminal, or the like through the output unit 705. Also, the separated audio stream is MPEGAU
The data is input to the DIO unit 706, and the reproduction processing including error detection as described above is performed. When an error is detected, the data adaptively corresponding to the flow of the signal is calculated as repair data, and this is output as interpolation data. The output audio data is passed through a D / A converter 707 to an audio output unit 7.
08 is output.

As described above, also in the video / audio reproducing apparatus,
If an error is detected in the audio data, the data that was in the signal flow adaptively is used as repair data,
Relatively easily and naturally in line with the signal flow,
Data can be repaired.

[0036]

As described above, according to the present invention, data can be easily restored by interpolating data calculated with reference to the flow of data signals before and after. Further, by sharing the buffer used at the time of restoration and the buffer used at the time of normal reproduction, memory resources can be saved.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an audio reproducing device according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram showing a structure of PCM data on a buffer in the audio reproducing device according to one embodiment of the present invention.

FIG. 3 is an explanatory diagram showing a first example of a method of calculating restoration data in the audio reproduction device according to the embodiment of the present invention.

FIG. 4 is an explanatory diagram showing a second example of a method of calculating restoration data in the audio reproduction device according to the embodiment of the present invention.

FIG. 5 is an explanatory diagram showing a third example of a method of calculating restoration data in the audio reproduction device according to the embodiment of the present invention.

FIG. 6 is an explanatory diagram showing a fourth example of a method of calculating restoration data in the audio reproduction device according to the embodiment of the present invention.

FIG. 7 is a block diagram showing a configuration of a video / audio reproduction device incorporating the audio reproduction device according to one embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 100 input unit 101 bit stream decomposition unit 102 error detection unit 103 scale factor decoding unit 104 bit allocation decoding unit 105 inverse quantization unit 106 subband synthesis unit 107 control unit 108 buffer unit 109 output unit 200 to 204 memory area 700 recording medium 701 Separation unit 702 MPEGCODEC unit 703 Format conversion unit 704 D / A conversion unit 705 Output unit 706 MPEGAUDIO unit 707 D / A conversion unit 708 Audio output unit

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (reference) // H04N 7/24 H04N 7/13 A F term (reference) 5C059 MA00 RF01 RF04 SS12 SS13 SS14 TA76 TC22 UA05 5D045 DA20 5J064 BA16 BB04 BB08 BC01 BC07 BD02 BD03 5K014 AA01 BA06 FA06 5K028 AA01 EE03 EE08 KK01 KK03 PP12 RR04 SS05 SS15 SS24

Claims (13)

[Claims] 1. A sound reproducing apparatus for reproducing sound by decompressing data compressed using sub-band coding and restoring a sound signal, wherein the sub-band coded data is separated into a desired number of bits. Separating means for performing error detection of data separated by the separating means, and control means for performing control for adaptively restoring audio output based on a signal from the error detecting means. An audio reproduction device characterized by the above-mentioned. 2. A sound reproducing apparatus for reproducing sound by decompressing data compressed by using sub-band coding and restoring a sound signal, comprising the steps of: Separating means for separating an error detection word, scale factor selection information, a scale factor, and a sample; an error detecting means for detecting an error of data separated by the separating means; and decoding the scale factor separated by the separating means. Scale factor decoding means, bit allocation decoding means for decoding the bit allocation separated by the separation means, inverse quantization for performing inverse quantization by the separated samples and the bit allocation decoded by the bit allocation decoding means Means, and a sub calculated by the inverse quantization means. Sub-band combining means for combining band data and outputting PCM data; buffer means for storing the PCM data;
Control means for adaptively restoring and outputting the M data;
An audio playback device, comprising: 3. The audio reproducing apparatus according to claim 2, wherein the error detection means performs error detection on a part of the header, the bit allocation, and the scale factor selection information. 4. The audio reproducing apparatus according to claim 2, wherein the buffer means has a capacity of at least a predetermined number of coding units. 5. The audio reproducing apparatus according to claim 4, wherein the predetermined number is at least five. 6. The control means according to claim 2, wherein said control means, upon receiving said error detection signal, obtains a predetermined number of interpolation data from predetermined data under predetermined conditions and based on said data, An audio reproducing device for restoring data. 7. The data processing apparatus according to claim 2, wherein the control unit, upon receiving the error detection signal, sets two data before and after the error-detected data from two data before and after the error-detected data under a predetermined condition. Speech reproduction characterized in that a predetermined number of interpolation data to be interpolated between the last two data is obtained, and the data between the previous two data and the rear two data is restored based on the data. apparatus. 8. The condition according to claim 7, wherein the predetermined condition is that the polarity of the rate of increase of the signal level of the two preceding data is opposite to the polarity of the rate of increase of the signal level of the two subsequent data. The control means calculates an extrapolation point of the preceding two data and an extrapolation point of the following two data so as to be at the same time, and calculates the extrapolation point of the preceding two data and the following two data. In the case where the arrangement of the preceding two data and the following two data among the obtained extrapolation points is convex upward, the smaller absolute value of the signal level is determined by comparing the absolute value of the signal level between the preceding two data and the following two data. If the data is selected as interpolation data to be interpolated between the two preceding data and the following two data are convex downward, the one having the larger absolute value of the signal level is regarded as the preceding two data. Select as interpolation data to be interpolated between the last two data Audio reproduction device, characterized in that. 9. The control unit according to claim 8, wherein when there are a plurality of extrapolated points obtained from the two data before and two extrapolated points obtained from the two data after the data, the control means: Interpolation between two extrapolated points determined from two data before and two extrapolated data at the same point in time, and interpolated between two data before and two data after An audio reproducing apparatus characterized by selecting data for use. 10. The condition according to claim 7, wherein the predetermined condition is that the polarity of the rate of increase in the signal level of the two preceding data is opposite to the polarity of the rate of increase in the signal level of the two subsequent data. The control means calculates an extrapolation point of the preceding two data and an extrapolation point of the following two data so as to be at the same time, and calculates the extrapolation point of the preceding two data and the following two data. In the case where the arrangement of the preceding two data and the following two data among the obtained extrapolation points is convex upward, the one having the smaller absolute value of the signal level is selected as the first data, In the case where the sequence of the last two data is convex downward, the data having the larger absolute value of the signal level is selected as the first data, and the data one by one before and after the data in which the error is detected are selected. , Linear at the same time as the extrapolated point The second data to calculate the interpolation data, linear interpolation data calculated from the first data and the second data,
An audio reproducing apparatus characterized in that the data is interpolation data to be interpolated between two data before and two data after. 11. The apparatus according to claim 10, wherein, when a plurality of the first data and a plurality of the second data are present, the control means controls the first data and the second data at the same time. An audio reproducing apparatus characterized in that each linear interpolation data calculated from data is used as interpolation data to be interpolated between two data before and two data after. 12. The predetermined condition according to claim 6, wherein the predetermined condition is that the polarity of the rate of increase in the signal level of the preceding two data is the same as the rate of the rate of increase of the signal level of the following two data. The control means includes: a first and a second data before and after the error-detected data;
An audio reproducing apparatus characterized in that linear interpolation data calculated from each data is used as interpolation data to be interpolated between data before and after. 13. A recording medium having the audio reproducing apparatus according to claim 1 for recording a compressed system stream, and a video stream for recording the recorded system stream. Separation means for separating the video stream into audio streams; video decoding means for decoding the video stream; format conversion means for format-converting the video signal decoded by the video decoding means; and an analog signal converted by the format conversion. Video D / A conversion means for converting the audio stream into an audio output, audio decoding means for decoding the audio stream, and audio D / A conversion means for converting a signal output from the audio decoding means into an analog output. A video / audio reproduction device characterized by the following.