JP3226711B2 - Compressed information reproducing apparatus and compressed information reproducing method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a compressed information reproducing apparatus for reproducing compressed information such as MPEG audio data and the like .
And a compressed information reproducing method .

[0002]

2. Description of the Related Art In recent years, an information reproducing apparatus for reproducing information obtained by digitizing data such as video and audio has been developed, and is widely used, for example, as a reproducing apparatus for movie software or karaoke. Recently, the data compression method for data such as video / audio has been changed to MPEG (Moving P
(icture Image Coding Expert Group) has already been internationally standardized. The MPEG method is a method of performing variable length compression on video data, audio data, and the like.

As shown in FIG. 5, an information reproducing apparatus (compressed information reproducing apparatus) of the MPEG system generates a compressed audio (MPEG audio data) by an encoder 51 in response to an input of an audio, and transmits the compressed audio (MPEG audio data) to a communication path or a storage medium. And the MPEG audio data is input to the decoder 52 via the.

When this compressed information reproducing apparatus is applied to, for example, fields such as broadcasting and communication, data is always transmitted between the encoder 51 and the decoder 52 at a constant rate. , The maximum processing speed of the decoder 52 is
Only the processing capability determined by the sampling frequency and the bit rate needs to be provided.

[0005] By the way, it is conceivable that this compressed information reproducing apparatus will be applied to fields such as storage media, that is, VTRs and optical disks in the future. In this case,
Since special playback such as reverse playback, fast forward / slow playback, and the like are frequently used, it is inevitable to reproduce at a desired speed, particularly after transmitting audio simultaneously with video.

In this case, it is conceivable that the video and the audio are simultaneously reproduced at a double speed or a triple speed by performing the data thinning.

However, data compressed by the MPEG method requires data continuity for reproduction, and therefore is basically suitable for special reproduction for thinning out data, ie, double speed reproduction or triple speed reproduction. Not.

Therefore, in order to realize this, it is necessary to devise a data structure or add a circuit.

For example, high-speed reproduction of video / audio data or the like compressed by the MPEG method using a VTR (normally,
In this case, the simplest method is to increase the transfer rate of the data input to the decoder 52, thereby improving the processing speed of the decoder 52 and instantaneously decode the data. Then, the data may be thinned out for each audio sample.

In order to make hardware compatible with this method, the processing speed of the decoder 52 must be dramatically improved. That is, in addition to the above-described encoding bit rate and sampling frequency, the decoder 52 is required to have a processing speed several times as high as the reproduction speed.

However, at this time, it is almost impossible to improve the processing speed of the decoder 52, and it can be said that the feasibility is extremely low.

[0012]

As described above, in the above-mentioned conventional compressed information reproducing apparatus, when the compressed information is specially reproduced, that is, when the compressed sound or the like is reproduced at a speed several times higher than the normal speed, the processing of the decoder is performed. Although it is necessary to dramatically increase the speed, at present, it is extremely difficult to significantly increase the processing speed of the decoder in terms of hardware, and there has been a problem that it is not possible to adapt to such special reproduction.

The present invention has been made to solve such a problem, and an object of the present invention is to provide a compressed information reproducing apparatus which can adapt to special reproduction of compressed information without improving the processing speed of a decoder. .

[0014]

Compression information reproducing apparatus according to claim 1 SUMMARY OF THE INVENTION In order to achieve the above object, the analog
The audio signal is divided into frames and compressed and encoded.
The bit rate information of the frame is set in the header of each frame.
Compressed information reproducing apparatus for decoding / reproducing compressed data
In a storage means for storing the compressed data, bit rate from the contents of the header of each frame of the <br/> compressed data
A header analysis unit for extracting a frame, and an optimal decimation frame from the bit rate extracted by the header analysis unit.
The number of frames is determined, and each frame in the compressed data stored in the storage means is selected at predetermined intervals in accordance with the optimum number of thinned frames and a reproduction speed designated in advance, and a frame selection to be read at a standard rate is performed. Means, and decoding means for decoding and reproducing the compressed data read at a standard rate by the frame selecting means.

The compressed information reproducing apparatus according to the second aspect, Anna
The log audio signal is divided into frames and compressed and encoded.
And the frame bit rate information in the header of each frame.
Information reproduction for decoding / reproducing compressed data provided with
In the apparatus, storage means for storing the compressed data,
From the contents of the header for each frame of the compressed data,
A header solution that extracts the pulling frequency and bit rate
Analysis section and the sun extracted by the header analysis means.
Optimum decimation filter based on coupling frequency and bit rate
Calculate the number of frames and specify the optimal number of thinned frames in advance.
Stored in the storage means in accordance with the playback speed thus set.
Select each frame in the compressed data at predetermined intervals
Frame selection means for reading at a standard rate, and
The pressure read by the frame selection means at a standard rate.
Decoding means for decoding and reproducing the compressed data
It is characterized by.

[0016] compression information reproducing apparatus according to claim 3, in the compression information reproducing apparatus according to claim 1 or 2, wherein said frame selection means, the reproduction speed N, when the number of the optimal frame decimation was X, Frame read interval is N
It is characterized in that every X frames are read, X frames are continuously read out, and the data of (N-1) × X frames following this are thinned out .

[0017] The compressed information reproducing apparatus according to claim 4, in the compression information reproducing apparatus according to claim 1 or 2, wherein the off
The frame selection means uses the optimal thinning-out function according to the bit rate.
A table in which frame number information is stored in advance is provided.
It is characterized in that the optimum number of thinned frames is obtained from the table .

According to a fifth aspect of the present invention, there is provided the compressed information reproducing apparatus according to the fourth aspect, wherein
Indicates that the optimal number of thinned frames according to the bit rate is 1
Total data of frame bit amount and optimal number of thinned frames
Is set to be close to the preset value.
It is characterized in that there.

The compression information reproducing method of claim 6, Anna
The log audio signal is divided into frames and compressed and encoded.
And the frame bit rate information in the header of each frame.
Information reproduction for decoding / reproducing compressed data provided with
The method comprises: temporarily storing the compressed data;
From the contents of the header for each frame of the compressed data,
And extract the optimal bit rate from the extracted bit rate.
The number of thinned frames is calculated, and the obtained optimum thinned frame is calculated.
According to the number and playback speed specified in advance.
Each frame in the compressed data is
Select and read at standard rate, read at this standard rate
Decoding and reproducing the compressed data thus output.
And The compressed information reproducing method according to claim 7,
The log audio signal is divided into frames and compressed and encoded.
And the frame bit rate information in the header of each frame.
Information reproduction for decoding / reproducing compressed data provided with
The method comprises: temporarily storing the compressed data;
Sampling from the contents of the header for each frame of the compressed data
The number of round trips and bit rate before extraction
Between the sampling frequency and bit rate
The number of thinned frames is calculated, and the obtained optimum thinned frame is calculated.
According to the number and playback speed specified in advance.
Each frame in the compressed data is
Select and read at standard rate, read at this standard rate
The compressed data thus output is decoded and reproduced .

[0020]

According to the first , second, sixth and seventh aspects of the present invention , the compression code
The encoded compressed data is stored in the storage means, and the bit rate is extracted from the contents of the header for each frame by the header analysis means, or the sampling frequency is
And the bit rate are extracted . Then, the extracted bit
Rate, or sampling frequency and bit rate
The optimal number of thinned frames is calculated from
In accordance with the reproduction speed designated in advance the number of can frame, frame selection means, each frame in the compressed data memory means read by the standard rate selected at predetermined intervals, to perform the thinning process of the compressed data . The compressed data for each frame read by the frame selecting means is decoded by the decoding means as usual, but the decoded reproduced sound is reproduced at the specified speed because the frames have already been thinned out. Is done.

That is, the provision of each unit having a simple configuration for thinning out data between the encoding unit and the decoding unit enables special reproduction of the compressed information without increasing the processing speed of the decoding unit. Become like

In the present invention , the head of each frame is detected from the compressed data, the header data following the head of the frame is analyzed, and the bit rate and sampling frequency of the compressed data are extracted as the contents of the header. Each frame in the compressed data is selected at predetermined intervals in accordance with the contents of these headers and a reproduction speed specified in advance, and is read out at a standard rate. Then, when the decoding means decodes the read compressed data as usual, the reproduced sound is reproduced at the specified speed because the frames have already been thinned out.

Therefore, the special reproduction of the compressed information can be realized without increasing the processing speed of the decoding means.

According to a third aspect of the present invention, the predetermined reproduction speed is set to N and the optimum number of thinned frames is set to X.
In this case, the frame reading interval is set to every N × X frames, and after reading X frames continuously, data for the following (N−1) × X frames is discarded.

For example, when the optimum number of thinned frames is 7 and the reproduction speed is twice, the frame reading interval is every 14 frames. After reading 7 frames continuously, the data of the following 7 frames is discarded. Become like

As a result, even when the compressed data obtained by compressing and encoding the analog signal is transferred, the decoder can reproduce the analog signal at the designated speed.

According to the fourth aspect of the present invention , the frame selecting means
The column shows the optimal number of thinned frames according to the bit rate.
Is provided with a table stored in advance.
Find the number of properly thinned frames. Therefore, read
When the decoding means decodes the compressed data normally,
Frames are decimated by the appropriate number of decimated frames,
The playback sound will be played at the speed specified above.
You .

According to the fifth aspect of the present invention, there is provided the fourth aspect.
Table shows the optimal thinning frame according to the bit rate.
The number of frames is the bit amount of one frame and the optimal thinning frame
The total amount of data will be close to the preset value
Is set so that the quality of the playback sound and the
Improves balance and makes high-speed playback easier to hear
It will be something. By setting the setting value to 1k bytes,
The most sensible human sound quality of reproduced sound and confirmation of information
High-speed playback sound is easy to hear.
Become

In the present invention, since the analog signal to be compression-encoded is audio information, high-speed reproduction of compressed audio can be realized.

[0030]

Embodiments of the present invention will be described below in detail with reference to the drawings.

FIG. 1 is a diagram showing a configuration of an encoder of a compressed information reproducing apparatus according to one embodiment of the present invention, and FIG. 2 is a diagram showing a configuration of a preprocessor and a decoder of the compressed information reproducing apparatus. .

As shown in FIG. 1, the encoder 1 comprises a sub-band analysis filter bank 2, a scale factor selection information section 3, a scale factor extraction section 4, a fast Fourier transform section (FFT) 5, and a psychoacoustic model 6. , A dynamic bit allocation unit 7, a side information encoding unit 8, a linear quantization unit 9, a bit compression unit 10, a bit stream forming unit 11, and the like. The encoder 1 is connected to a preprocessor 20 via a storage medium or the like.

As shown in FIG. 2, the preprocessor 20 includes a dual port memory 21, a SYNC detection unit (synchronization detection unit) 22, a header analysis unit 23, an audio frame selection control unit 24, a write address generation unit 25, It comprises a read address generation unit 26, a clock reproduction unit 27, a storage unit 28, and the like. The storage unit 28 stores the conversion table 2
9 is stored. Dual port memory 21
It has two address ports for input and output,
The input voice frame is stored. SYNC detector 2
2 is the SYN at the head of the input audio frame
The C word is detected and the beginning of the frame is confirmed. The header analysis unit 23 analyzes the header including the SYNC word from the head of the frame confirmed by the SYNC detection unit 22, and extracts various information such as a bit rate and a sampling frequency (fs). The audio frame selection control unit 24 receives the specification information such as the input bit rate and the sampling frequency (fs) and the reproduction speed designation from an external device (not shown) (not shown), and receives the write address generation unit 25 and the read address generation unit. 26 to execute a data thinning process.

The decoder 30 comprises a clock circuit 31, a bit stream decomposition / error detection section 32, a side information decoding section 33, an inverse quantization section 34, a subband synthesis filter bank 35, and the like.

Normally, MPEG audio data (MPEG audio data) generated by the encoder 1 is provided with a header for each fixed amount of stream.
In this header, a SYNC word (synchronization information) is inserted, followed by specification information such as a bit rate and a sampling frequency (fs). The amount of data allocated to this header is 4 bytes including the SYNC word. In the decoder 30, the decoding processing of the encoded audio data becomes possible only after a data stream sandwiched between two SYNC words is obtained.

The interval between SYNC words is called an audio frame. The data amount (N) constituting this audio frame is defined by the following equation.

N = 12 × (bitrate) / (sampling frequency) Expression (1) [Layer1] N = 144 × (bitrate) / (sampling frequency) Expression (2) [Layer 2 and 3] Note that the sampling frequency is 44.1. Although N is not an integer only in the case of kHz, the above relationship is satisfied by averaging by processing (indispensable processing of the encoder) for adjusting the number of bytes constituting several consecutive frames (for example, 49 frames).

The storage unit 28 stores a conversion table 29 into which the value of N expressed in the above equations (1) and (2) is substituted. In this embodiment, the conversion table 29 of [Layer 2and 3] is used. The conversion table 29 created based on (2) will be described.

As shown in FIG. 3, the conversion table 29
Shows the bit rate N on the vertical axis, the bit amount N in one audio frame for each sampling frequency (32 kHz, 44.1 kHz, 48 kHz, etc.) on the horizontal axis, and the optimal number of thinned frames X for the bit amount, using the encoding parameter. Is set as
The optimum number of thinned frames X is the total data amount (N × X)
Is set to be close to 1K bytes. Therefore, the numerical value of the optimal thinned frame number X is small for the high bit rate, and the numerical value of the optimal thinned frame number X is large for the low bit rate.

Normally, when the basic unit of processing is increased, the quality of the reproduced voice is improved, but the reproduction time of the data discarded by the thinning-out processing is also increased, so that the confirmation of information is inferior. This is most noticeable in the case of

On the other hand, when the basic unit of processing is reduced, the recognizability of information is relatively improved, but the quality of reproduced sound is deteriorated. In particular, when music or the like is reproduced when the coding bit rate is low, the deterioration of the sound quality is noticeable. Therefore, a setting value of 1 Kbyte was set as a setting value that appropriately satisfies the above two conflicting conditions. When the data volume of the double-speed processing is set with this 1 Kbyte, the quality of the reproduced sound and the confirmability of the information are balanced, and the listening is easy.

In the compressed information reproducing apparatus, the encoded audio stream (encoded audio data) is transmitted at a required data rate from the encoder 1 via the storage medium even during double-speed reproduction.

This audio stream is first sent to the preprocessor 2
0 dual port memory 21 and SYNC detector 22
Entered as The dual port memory 21 stores an audio stream in which addresses are assigned in frame units.

When a voice frame is input to the SYNC detection unit 22, the SYNC word at the head of the frame is detected, and the head of the frame is confirmed.

When the head of the frame is confirmed, the contents of the header starting from the head of the frame are analyzed by the header analysis unit 23 at the subsequent stage of the SYNC detection unit 22, and the sampling frequency (fs) and the encoding bit rate are analyzed by this analysis. Are extracted and output to the audio frame selection control unit 24.

On the other hand, the audio frame selection control unit 24
The double speed number of the sound to be reproduced in advance is specified from the outside (microcomputer or the like).

Therefore, the audio frame selection control unit 24
Determines the interval of the thinning process by referring to the conversion table 29 based on the sampling frequency (fs), the encoding bit rate, and the double speed number.

Here, in order to obtain the optimum thinned-out frame X and the reproduction sound at N times speed, the sound frame selection control unit 24
Sets the interval of the thinning process to N × X frames by referring to the conversion table 29 based on the obtained parameters, and continuously outputs X frames to the decoder 30 at the interval, and then (N−1) ) Thinning out (discard) data for X frames
The read address generator 26 is instructed to generate a read address, and only the audio frame of the corresponding address is read from the dual port memory 21.

For example, the sampling frequency (fs) is 32 kHz
If the encoding bit rate is 32 kBit / s and the data amount N of the audio frame is 144 kbps, the optimum number of thinned frames X is determined to be "7" based on the conversion table 29 (the top row of the conversion table 29). And if the double speed number is twice as high, 7 frames are played back and the next 7 frames are thinned out.
In this way, the sound is reproduced at double speed. If the double speed is 3 times, 7 frames are reproduced and the next 14 frames are thinned out. As a result, the sound is reproduced at 3 times speed.

That is, the pre-processor 2 in the preceding stage of the decoder 30
0, a read address is generated by referring to the conversion table 29 based on the parameter obtained from the header and the double speed number input from the outside, and the data of the dual port memory 21 is read out by thinning it out. A predetermined number of audio frame data are thinned out in units, and the data is input to the decoder 30 at exactly the same data rate (standard rate) as during standard reproduction. When the coded audio data after the thinning processing is input to the decoder 30, the decoder 30 executes a decoding process that is not different from a normal one. As a result of this processing, sound can be reproduced at the specified speed.

As described above, according to the present embodiment, the encoder 1
And the decoder 30 is used as it is, and the preprocessor 20 having a simple configuration such as the dual-port memory 21 and the voice frame selection control unit 24 is added (before the decoder 30) in the meantime. High-speed playback sound can be obtained. In particular, as the performance of the decoder 30, it is sufficient that the decoder 30 has a conventional maximum processing speed for executing the standard reproduction determined by the encoding bit rate and the sampling frequency. It is possible to immediately respond to double-speed reproduction of MPEG audio.

Further, by setting the number of audio frames, which is the basic unit of the double-speed processing, closer to 1 kbyte, the quality of the reproduced sound and the confirmation of the information can be balanced, so that the audio of the double-speed reproduction can be easily heard. can do.

Next, another embodiment of the present invention will be described with reference to FIG.

FIG. 4 is a view showing a modification of the above embodiment. The same components as those in the above embodiment are denoted by the same reference numerals, and description thereof will be omitted.

In the above embodiment, the dual port memory 21 is used for data rate conversion during double-speed reproduction, and the preprocessor 20 has a control master for data transfer from the preprocessor 20 to the decoder 30. In this embodiment, however, as shown in the figure, a FIFO (First-In-First-Out) memory 4 is used to perform data rate conversion during double-speed reproduction.
I use 1. Further, a control master for setting the timing of data transfer from the preprocessor 20 to the decoder 30, that is, a clock recovery unit 27 is provided on the decoder 30 side.

In this case, the data thinning process is executed by controlling the writing of data into the FIFO memory 41 by the audio frame selection control unit 24.

At this time, the preprocessor 20 generates a write pulse associated with the thinning process. FIFO memory 4
The data reading from 1 is executed using the data transfer clock of the clock recovery unit 27 generated from the decoder 30 side as a reading pulse. According to this embodiment, by providing the clock recovery unit 27 on the decoder 30 side, the functions of the preprocessor 20 and the decoder 30 can be differentiated and the versatility can be increased, so that double-speed reproduction can be performed more efficiently. Sound can be obtained.

Further, the FIFO memory 4
1 and the addition of the pre-processing unit 20 having a simple configuration such as the audio frame selection control unit 24, so that a good double-speed reproduced sound can be obtained without a complicated double-speed processing unit or the like at the subsequent stage of the decoder 30. Therefore, it is possible to obtain the effects of reducing the current hardware scale and simplifying the control routine.

[0059]

According to the onset bright As described in the foregoing, according to the content and predesignated playback speed of the header obtained from the compressed data from the encoding means, the compressed frame selection means of the storage means By selecting each frame in the data at predetermined intervals and reading it out at a standard rate, the compressed data is thinned out, so that the read out compressed data is input to the decoding means and decoded as before, The decoded reproduced sound is reproduced at the specified speed.

That is, by providing a storage unit or a frame selection unit having a simple configuration for thinning out data between the encoding unit and the decoding unit, the compression information can be compressed without increasing the processing speed of the decoding unit. Special playback can be realized.

By setting the thinning interval to be closer to 1 kbyte, the quality of the reproduced sound and the confirmability of the information can be balanced, and the high-speed reproduced sound can be easily heard.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a compressed information reproducing apparatus according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a main configuration of the compressed information reproducing apparatus of FIG. 1;

FIG. 3 is a diagram showing a conversion table of the compressed information reproducing device.

FIG. 4 is a block diagram showing a double-speed playback compatible player according to another embodiment.

FIG. 5 is a diagram schematically showing a configuration of a conventional compressed information reproducing apparatus.

[Explanation of symbols]

1. Encoder, 2. Sub-band analysis filter bank, 3.
... Scale factor selection information section, 4 ... Scale factor extraction section, 5 ... Fast Fourier transform section (FFT), 6 ... Psychological auditory model, 7 ... Dynamic bit allocation section, 8 ... Side information coding section, 9 ... Linear quantization Section, 10 ... bit compression section, 11 ...
Bit stream forming unit, 20: preprocessor, 21: dual port memory, 22: SYNC detecting unit, 23: header analyzing unit, 24: audio frame selection control unit, 25: write address generating unit, 26: read address generating Unit, 27: clock reproducing unit, 28: storage unit, 29: conversion table, 30: decoder, 31: clock circuit, 32 ...
Bit stream decomposition / error detection unit, 33: side information decoding unit, 34: inverse quantization unit, 35: subband synthesis filter bank.

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H03M 7/30 G10L 11/00 H04N 5/93

Claims (7)

(57) [Claims] 1. An analog audio signal is divided into frame units.
Is compressed and encoded, and the frame is added to the header of each frame.
Of compressed data provided with the bit rate information of
In the compression information reproducing apparatus for raw, storage means for storing the compressed data, bits from the contents of the header of each frame of the compressed data
A header analyzing unit for extracting a rate, and the bit rate extracted by the header analyzing means.
The optimal number of thinned frames is calculated from
In accordance with the reproduction speed designated in advance with the over arm number, each frame in said compressed data stored in the storage means selected at predetermined intervals, and the frame selecting means for reading a standard rate, the frame selecting means Decoding means for decoding and reproducing the compressed data read at a standard rate according to (1). 2. An analog audio signal is divided into frame units.
Is compressed and encoded, and the frame is added to the header of each frame.
Of compressed data provided with the bit rate information of
A compressed information reproducing apparatus that generates the compressed data, and a storage unit that stores the compressed data;
A header solution that extracts the pulling frequency and bit rate
And analyzing unit, the sampling extracted by the header analyzer
Optimal number of thinned frames from frequency and bit rate
And determine the optimal number of thinned frames and the
The pressure stored in the storage means according to the raw speed.
Each frame in the compressed data is selected at predetermined intervals, and
Frame selecting means for reading at a standard rate, and the frame reading means for reading at a standard rate.
Decoding means for decoding and reproducing the compressed data.
Compression information reproducing apparatus, characterized in that the. 3. A compressed information reproducing apparatus according to claim 1 or 2, wherein said frame selection means, the reproduction speed N, the optimum if the decimation number of frames was used as X, every N × X frames frame readout interval A compressed information reproducing apparatus characterized in that after continuously reading X frames, data of (N−1) × X frames following this is thinned out . 4. A compressed information reproducing apparatus according to claim 1 or 2, wherein said frame selection means, the optimal frame decimation number information corresponding to the bit rate in advance
It has a table that is stored, and
A compressed information reproducing apparatus for determining the number of frames to be compressed. 5. The compressed information reproducing apparatus according to claim 4, wherein said table has an optimum thinning-out function corresponding to a bit rate.
The number of frames is equal to the bit amount of one frame and the optimal thinning frame.
The total data volume of the number of programs approaches the preset value
A compressed information reproducing apparatus characterized by being set as follows . 6. An analog audio signal is divided into frame units.
Is compressed and encoded, and the frame is added to the header of each frame.
Of compressed data provided with the bit rate information of
In the method for reproducing compressed information, the compressed data is temporarily stored, and a bit is determined from the content of a header for each frame of the compressed data.
Extract the trait and extract the optimal number of thinned frames from the extracted bit rate.
The calculated, pre specified this determined optimal thinning the number of frames reproduced
The compressed data temporarily stored according to the speed
Select each frame at regular intervals and read at standard rate.
And decodes the compressed data read at this standard rate.
Compression information reproducing method characterized by reproducing Te. 7. A compressed information reproducing method for decoding / reproducing compressed data in which an analog audio signal is divided into frame units and compression-encoded, and wherein the header of each frame is provided with bit rate information of the frame, Is temporarily stored, a sampling frequency and a bit rate are extracted from the contents of the header of each frame of the compressed data, and an optimum number of thinned frames is obtained from the extracted sampling frequency and bit rate. Each frame in the compressed data that is temporarily stored is selected at predetermined intervals and read at a standard rate in accordance with the optimum number of thinned frames and a reproduction speed specified in advance, and read at the standard rate. And decoding and reproducing the compressed data.