JP3916615B2 - Multiple encoded signal reproduction apparatus - Google Patents

Description

  The present invention relates to an encoded signal reproducing apparatus, and in particular, simplification of the circuit configuration of an apparatus that reproduces a digital code string in which audio, video, and other additional information are multiplexed, such as video CD, DVD, and digital CS broadcasting. It is about what aimed at.

  In recent years, as seen in video CDs, DVDs, digital CS broadcasts, and the like, media that digitally encode and multiplex and record and transmit additional information signals such as image signals, audio signals, and caption information have rapidly spread. It's getting on. However, it is essential to reduce the cost of the reproduction apparatus in order to spread the reproduction apparatus for the encoded and multiplexed digital signal as described above to homes and the like. For this purpose, it is required to implement a separation device for separating multiplexed signals and a decoding device for decoding each separated digital signal with a simple and compact circuit configuration.

  By the way, in such digital media, the encoding method defined by the MPEG standard is generally used for video signals in many cases. On the other hand, a coding method other than the MPEG coding method is considerably adopted for the audio signal. The video and audio encoded data is multiplexed by a multiplexing method defined by the MPEG system standard.

  FIG. 3 shows an example of data obtained by encoding video information, audio information, and additional information multiplexed by a packet. The multiplexed and encoded data is byte aligned. Each of the audio signal and the video signal is digitally encoded by an encoding device and multiplexed by a packet unit by a multiplexing device. At the beginning of the packet, there is a synchronization signal S indicating the beginning of the packet (hereinafter referred to as a packet start code prefix) S and an identifier I for distinguishing whether the packet corresponds to a voice, image or additional information packet, A packet header H composed of packet length information L indicating the packet length, header length HL, video and audio synchronous reproduction information R, etc. is added. After the packet header H, one of video signal encoded data CDP, audio signal encoded data CDS, and additional information encoded data CDA is arranged according to the type of packet. The system layer is defined from the packet start code prefix to the header end, and the encoded data portion of video, audio, and additional information data after the header is defined as an elementary layer.

  In encoding a video signal, hierarchical encoding of the video signal is performed, and a code string indicating the start of the hierarchy and a code string indicating the hierarchy name are used. The same code string is used for the code string indicating the start of this hierarchy and the start code prefix code (prefix) indicating the head of the packet.

  Usually, since the prefix code of the packet start code appears in the data string for each data unit indicated by the packet length included in the packet header information, the confusion between the packet start code prefix and the hierarchical start code of the encoded video data is Does not happen.

  However, in a packet including video encoded data, unlike a packet including audio encoded data, multiplexed data having an indefinite packet length may be input. Such a situation occurs, for example, in a multiplexed data sequence obtained by converting a transport packet used in digital CS broadcasting or the like into a PES (Packetized Elementary Stream) packet. A conventional method for separating multiplexed data when the packet length of the encoded video data is indefinite will be described below.

  FIG. 12 is a diagram illustrating an operation when reading out the encoded video data. The case where the packet boundary between the video packet 1 and the video packet 2 is located between ('x3') and ('00'). Show. The read pointer reads the video packet 1 as x0, x1, x2... And is sequentially transferred to the decode buffer. The pointer reaches “E0” of the video packet 2 beyond the packet boundary, and (“00”, “ When a pattern of “00”, “01”, “E0”) is detected, it is recognized for the first time that a packet boundary exists. At this time, the video buffer 2 (“E0”) is already stored in the decode buffer. ) Is transferred as valid data, the read pointer is returned to the packet boundary again, and the data corresponding to the pattern of ('00', '00', '01', 'E0') is Since the data is not originally the data to be decoded but is used in the system, the pointer is returned again so that this data is not sent to the decoding buffer.

  Consider two examples in FIG. 4 where the layer start code of the video encoded data is divided between two packets as a pattern in which the packet boundary is likely to be erroneous.

Hereinafter, the packet boundary determination operation will be described for each of the patterns in FIGS.
First, the pattern in FIG. 4A will be described. In the pattern of FIG. 4A, the layer start codes of the video encoded data are packet start codes ('00', '00) in ('00') and ('00', '01', '00'). ',' 01 ',' E0 ') and the packet header are divided into two packets.

  In such a case, the start code prefix code detection unit for detecting the start code prefix code reads data from the input buffer and detects the start code prefix code pattern. At this time, the read address of the input buffer is the address α. The start code pre-code detection unit detects the pattern ('00', '00', '01') and then informs the start code determination unit located in the subsequent stage that the pattern has been detected, thereby determining the start code. Start the department.

  Next, the start code determination unit reads the data at the address β and determines whether this is an identifier ('B9' to 'FF') indicating the beginning of the packet or a hierarchical start code ('00' to '00' to 'B8').

  In the pattern of FIG. 4A, since the data ('E0') at the address β is an identifier indicating the start of the packet, the start code determination unit notifies the header analysis unit that the packet head has been detected, and the header Start the analysis unit.

  At this time, since the data at the address γ is a part of the encoded video data, it is necessary to connect to the subsequent part of the subsequent video packet data and transfer it to the decode buffer. Therefore, before starting the header analysis unit, in order to transfer the data of the address γ, the read address of the input buffer is reset from the address β to the address γ that is an address before it, and the address γ It is necessary to perform the operation of transferring the data to the decode buffer. In addition to performing data transfer at address γ, it is also necessary to store information that the data at address γ is part of the hierarchical start code of the video encoded data.

  In the case of FIG. 4B, after the start code discriminating unit discriminates the identifier indicating the start of the packet at the address β, the data at the addresses γ and δ, which are part of the video encoded data, is transferred to the decode buffer. In addition, it is necessary to perform control for returning the read address of the input buffer corresponding to the address β to the address γ and transferring the data of the addresses γ and δ to the decode buffer, which is more than the case shown in FIG. You will move the pointer a lot. Further, as in the case of FIG. 4A, it is necessary to store information that the data of the addresses γ and δ is a part of the layer start code of the video encoded data. Information that the stored data ('00') at address γ may be part of the hierarchical start code is used to detect the start code of the video playback unit in the encoded data portion of the next video packet. .

  As described above, since the video packet has an indefinite length, the conventional encoded signal reproduction apparatus performs complicated control such as not only advancing the read pointer of the input buffer but also returning it at the packet boundary. ing.

  That is, in this conventional encoded signal reproduction device, data transfer and discrimination are simultaneously performed using the same input buffer read pointer, so that the input data can be discriminated as system layer data. It is necessary to advance the write pointer of the decode buffer to perform extra writing. For this reason, the work of advancing the write pointer of the decode buffer is performed by the decode buffer write pointer correcting unit. In addition, if the input data is system layer data, it is necessary to detect the start code again. For this reason, the value of the pointer is temporarily returned, but this prevents the input data from being overwritten. A buffer read pointer correction unit is provided to correct the pointer, and an input buffer protection unit is provided to store input data so that data input to the input buffer is not overwritten, thereby protecting the input data. I have to. For this reason, the configuration and control of the apparatus are complicated.

  Conventional encoded signal reproduction apparatuses are configured as described above, and video and audio signal reproduction apparatuses often adopt a pipeline structure. In the pipeline, the data bus width is specified, and the encoded data is transferred and decoded with the specified bus width. However, the final portion of the encoded data may be less than the data bus width. In order to control the transfer of data less than this data bus width, it is necessary to perform data transfer control different from normal data transfer, that is, in a processing pipeline for enabling access in units of 1 byte. However, there was a problem that hardware was complicated accordingly.

  The present invention has been made to solve the above-described problems of the prior art, and it is an object of the present invention to obtain an encoded signal reproducing apparatus capable of reproducing a digital signal with less hardware resources without requiring complicated control. Objective.

According to the encoded signal reproduction device of the present invention, the specific code sequence is inserted into the last packet of the encoded data regardless of whether or not the code sequence indicating the end of the encoded data is already included. Code string insertion means, code string indicating the end of the coded data from the code string of the coded data or / and termination code string detection means for detecting the specific code string, and coded data by the termination code string detection means When the code string indicating the end of the code or / and the specific code string is detected, the data bus width of the pipeline transfer including the end of the encoded data is the bus width of the pipeline transfer including the other data. And a formatter unit for adding a predetermined number of pseudo data to the tail of the code string indicating the end of the encoded data or the specific code string, and a pipeline transfer And it is obtained by so and a decoding buffer for sequentially storing the code sequence.
Further, according to the encoded signal reproduction apparatus according to the present invention, the input code string is a multiplexed encoded signal in which audio, video, and reproduction information associated therewith are multiplexed. It is.

Further, according to the encoded signal reproduction device according to the present invention, the specific code string insertion means is a host CPU, the specific code string is detected by the terminal code string detection means, and the predetermined number of code strings are detected. When the code string to which the pseudo data is added is transferred to the decode buffer, the host CPU is notified.

According to the encoded signal reproduction device of the present invention, the specific code sequence is inserted into the last packet of the encoded data regardless of whether or not the code sequence indicating the end of the encoded data is already included. Code string insertion means, code string indicating the end of the coded data from the code string of the coded data or / and termination code string detection means for detecting the specific code string, and coded data by the termination code string detection means When the code string indicating the end of the code or / and the specific code string is detected, the data bus width of the pipeline transfer including the end of the encoded data is the bus width of the pipeline transfer including the other data. And a formatter unit for adding a predetermined number of pseudo data to the tail of the code string indicating the end of the encoded data or the specific code string, and a pipeline transfer Because it was so and a decoding buffer for sequentially storing the code sequence, when there is no code sequence indicating the end of the encoded data is also an effect capable of performing reliably addition of pseudo data of the predetermined number.
Also, according to the encoded signal reproduction apparatus of the present invention, the input code string is a multiplexed encoded signal in which audio, video, and reproduction information associated therewith are multiplexed. Even when there is no code string indicating the end of the encoded data of the multiplexed encoded signal in which audio and video and reproduction information accompanying this are multiplexed, the predetermined number of pieces of pseudo data are surely added. There is an effect that can be done.

Further, according to the encoded signal reproduction device according to the present invention, the specific code string insertion means is a host CPU, the specific code string is detected by the terminal code string detection means, and the predetermined number of code strings are detected. Since the host CPU is notified when the code string to which pseudo data is added is transferred to the decode buffer, the host CPU detects that specific data has been input to the code buffer. This makes it possible for the host CPU to clear the pipeline and to prepare for the next encoding.

(Embodiment 1)
Hereinafter, an encoded signal reproduction apparatus according to Embodiment 1 of the present invention will be described with reference to the drawings. FIG. 9 is a block diagram showing a conceptual configuration of the encoded signal reproduction apparatus of the present invention. In FIG. 9, M1 receives the encoded signal, collates a predetermined code with the input code, The matching status information output means M2 for outputting information receives the matching status information. When ('00') appears in the code, in particular, it generates predetermined data and outputs it to the decoding buffer. Data formatting means. For data (excluding video signals) when ('00') does not appear in the encoded signal, the data is transferred to the decoding buffer without the data formatting means M2.

  FIG. 1 is a more detailed block diagram of the coded signal reproduction apparatus. This corresponds to the invention of claim 1 of the present application, and can be configured and controlled more easily than the conventional apparatus.

  In FIG. 1, reference numeral 2s1 denotes a multiplex encoded signal supply unit for supplying a multiplex encoded signal to the present multiplex encoded signal reproduction apparatus, up to the preceding stage of a decoder unit such as a video CD or DVD player or a digital CS tuner. The part corresponds to this. Reference numeral 2s2 denotes an input buffer that receives the output of the multiplex encoded signal supply unit 2s1, and includes, for example, a ring buffer. 2s4 gives a read control signal (SIG1) to the input buffer 2s2 to control the read address, thereby controlling the read buffer, SW1 is an output of the input buffer 2s2, and outputs it to three terminals a , B, c, and a switch for output, 2s3 receives a signal when the switch SW1 is switched to the terminal b, and detects a start code prefix (prefix code) included in the packet A detection unit detects a desired code sequence from a code sequence in which data is multiplexed using a synchronization signal composed of a start code prefix and a packet identifier and a packet length. 2s5 is activated in response to the status signal (SIG8) output from the start code state holding unit 2s14, and receives 1-byte start code value (SIG6) which is an output signal from the start code prefix detection unit 2s3. A start code discriminating unit for discriminating whether the code belongs to any packet of video signal, audio signal, or additional information, 2s6 is a header analysis start signal (SIG11) which is an output signal from this start code discriminating unit 2s5 And a header analysis unit that analyzes a header by receiving a signal from the input buffer 2s2 when the determination of the start code is completed and the switch SW1 is switched to the terminal a, and analyzes the code after the synchronization signal. Is.

  Also, 2s7 receives a reproduction signal (SIG4) which is an output signal of the header analysis unit 2s6 and holds reproduction information, and 2s14 is a status update signal (an output signal of the start code prefix detection unit 2s3). It is a start code state holding unit that receives SIG5) and holds the status updated by the signal. Reference numeral 2s13 denotes a reproduction information state signal (SIG17) that is an output signal of the header analysis unit 2s6, a formatter activation signal (SIG9) that is an output signal of the start code determination unit 2s5, and an output signal of the start code prefix detection unit 2s3. In response to the byte signal (SIG6), the reproduction information (PTS) (SIG7) that is the output signal of the reproduction information holding unit 2s7, and the status signal (SIG8) that is the output signal of the start code state holding unit 2s14, a specific numerical sequence is obtained. When it is input, it is a formatter that outputs a numerical sequence corresponding to this, and the information output by the matching state information output means comprising the start code prefix detection unit 2s3, the start code state holding unit 2s14, and the start code determination unit 2s5 Generate data based on video encoded data And performs insertion of data strings at predetermined positions.

  SW2 is a switch that selects and outputs one of the output of the c terminal of the start code prefix detection unit 2s3, formatter 2s13, and switch SW1, and 2s8 receives a notification signal (SIG20) output from the start code determination unit 2s5. , A data separation control unit that receives a signal from the input buffer 2s2 and recognizes a packet boundary when the switches SW1 and SW2 are switched to the terminals c and f, respectively, and performs data separation control. When finished, the start code determination unit 2s5 is notified using the notification signal (SIG20). 2s19 is a controller that receives the format start / end signal (SIG19) output from the formatter 2s13 and the header end signal (SIG18) output from the header analysis unit 2s6, and controls the switching of the switches SW1 and SW2. A switching control signal (SIG3, SIG10) for controlling the switch is output.

  2s is the above-described switch SW1, switch SW2, start code prefix detection unit 2s3, start code determination unit 2s5, header analysis unit 2s6, reproduction information holding unit 2s7, data separation control unit 2s8, formatter unit 2s13, start code state holding This is a multiple signal demultiplexing unit composed of the unit 2s14.

  Also, 2s10 is a decoding unit that extracts the reproduction information held in the reproduction information holding unit 2s7 as a host control signal C (SIG15) via the host bus (HB) and decodes the signal from the data separation control unit 2s8, 2s9 is a decoding buffer that receives and holds the output signal of the data separation control unit 2s8 and supplies it to the decoding unit 2s10, and has independent storage areas for video, audio, and additional information, and the above data The data output from the separation control unit 2s8 is allocated to a predetermined area and stored. 2s11 is a work memory used when the decoding unit 2s10 performs a decoding operation, 2s12 is a host CPU that exchanges a host control signal B (SIG14) for performing processing such as initial setting and resetting of the decoding unit 2s10, The host bus HB is connected to a reproduction information holding unit 2s7 and a decoding unit 2s10, and is configured to output a supply control signal (SIG12) for controlling the data supply operation of the multiplex encoded signal supply unit 2s1. Yes.

  With the above configuration, the multiplexed encoded signal supply unit 2s1, the host bus HB, the host CPU 2s12, the controller 2s19, the decode buffer 2s9, and the work memory 2s11, that is, the input buffer 2s2, the input buffer read control circuit 2s4, the switch SW1, A switch SW2, a start code prefix detection unit 2s3, a start code determination unit 2s5, a header analysis unit 2s6, a reproduction information holding unit 2s7, a data separation control unit 2s8, a formatter unit 2s13, a start code state holding unit 2s14, and a decoding unit 2s10. An encoded signal reproduction device 2 is configured.

  Further, in the above configuration, the decode buffer 2s9 and the work memory 2s11 are external memories realized by SDRAM or the like, and are generally configured by a chip different from an LSI that realizes other components.

  In the above configuration, the start code prefix detection unit 2s3, the start code determination unit 2s5, and the start code state holding unit 2s14 realize the coincidence state information output unit M1, and the formatter unit 2s13 replaces the data format unit M2. It is to be realized. Further, the header analysis unit 2s6 and the reproduction information holding unit 2s7 realize header analysis means.

  Next, the operation will be described. The multiplex encoded data string supplied from the multiplex encoded signal supply unit 2s1 is temporarily stored in the input buffer 2s2. At this time, the switch SW1 is first connected to the contact b under the control of the controller 2s19. The multiplexed data sequence stored in the input buffer 2s2 is output byte by byte toward the start code prefix detection unit 2s3 under the control of the input buffer read control circuit 2s4. The start code prefix detection unit 2s3, together with the start code state holding unit 2s14, detects a packet start code that is a head code of packetized data. Then, the start code determination unit 2s5 is activated using status information (SIG8) from the start code state holding unit 2s14 described later.

  The start code discriminating unit 2s5 uses which packet identifier is different depending on whether the packet is a video packet, an audio packet, or an additional information packet. If the packet identifier following the start code prefix code is an identifier representing a desired data string to be reproduced, the header analysis unit 2s6 is notified that the packet is a valid packet. The start code determination unit 2s5 activates the header analysis unit 2s6 regardless of whether the packet is a valid packet or an invalid packet.

  If it is determined by the start code determination unit 2s5 that the input packet is a voice packet or an additional information packet, the start code analysis is unnecessary, and therefore the start code determination unit 2s5 causes the switch SW1 to be contacted by the controller 2s19. At the same time, the switch SW2 is switched to the contact f, and the output packet of the input buffer 2s2 is directly output to the data separation control unit 2s8.

  The data separation control unit 2s8 controls data transfer of the voice packet or the additional information packet to the decode buffer 2s9.

  The decoding unit 2s10 decodes the audio packet or additional information packet stored in the decoding buffer 2s9 by an internal audio decoder or additional information decoder, temporarily stores the decoded signal in the work memory 2s11, and stores the decoded signal from the work memory 2s11. Read out and output to the outside as a playback signal output.

  Next, when the input packet is a video packet, the functions of the formatter 2s13 and the start code state holding unit 2s14 will be described with reference to the flowchart of FIG.

  As described above, the switch SW1 is first connected to the contact b (step S200) under the control of the controller 2s19, and the switch SW2 is not connected to any contact. The multiplexed data sequence stored in the input buffer 2s2 is output byte by byte toward the start code prefix detection unit 2s3 under the control of the input buffer read control circuit 2s4 (step S201). The coincidence state information output means M1 including the start code prefix detection unit 2s3 and the start code state holding unit 2s14 detects a packet start code prefix which is the head code of packetized data (step S202), and starts code The status holding unit 2s14 is notified that ('00', '00', '01') is input by the status update signal SIG5. Next, after starting the start code determination unit 2s5 using the status information (SIG8), the 1-byte start code prefix detection unit 2s3 reads from the input buffer 2s2, and notifies the start code determination unit 2s5 with the 1-byte signal SIG6 (step) S203).

  Next, in step S204, the start code determination unit 2s5 uses the fact that the packet identifier is different depending on whether the packet is a video packet, an audio packet, or an additional information packet. When the packet identifier following the start code prefix code is an identifier representing a desired data string to be reproduced, the header analyzer 2s6 determines that the packet is a valid packet. Notice. The start code determination unit 2s5 activates the header analysis unit 2s6 in step S205 regardless of whether the packet is a valid packet or an invalid packet. Whether the packet is a valid packet to be reproduced or an invalid packet that is not to be reproduced is determined by the host CPU by the SIG13 and SIG14 via the host bus (HB) at the start of reproduction. This is done by collating with the identifier.

  When it is determined by the determination result of the start code determination unit 2s5 that the input packet is a video packet, the controller 2s19 switches the switch SW1 from the contact b to the contact a (step S206), and the video is transmitted to the header analysis unit 2s6. Output the packet. At this time, the switch SW2 is not connected to any contact.

  The header analysis unit 2s6 receives the packet from the input buffer 2s2, analyzes the packet length included in the packet header, the reproduction information used at the time of reproduction (step S207), and uses the reproduction information as the reproduction information holding unit. It is stored and held in 2s7 (step S208). The reproduction information includes information called PTS for synchronizing the reproduction time of audio and video, and information such as a flag indicating the presence / absence information of the PTS for each reproduction basic unit data included in each packet. It is.

  Further, the header analysis unit 2s6 determines the end of the header based on the header information (step S209). While executing the processing of these video data, the controller 2s19 controls the switch SW2 so that it is not connected to any contact, and activates the data separation control unit 2s8 and the start code prefix detection unit 2s3 (steps). S210). The data separation control unit 2s8 controls data transfer to the decode buffer 2s9 based on the data separation information held by the header analysis unit 2s6.

  The encoded video data stored in the decode buffer 2s9 is decoded by the video decoder in the decoding unit 2s10 as in the case of encoded data of audio and additional information, and a reproduction signal is output. The decoding unit 2s10 performs a decoding process using the work memory 2s11. At this time, the decoding operation is performed so that the reproduction information held in the reproduction information holding unit 2s7 is synchronized with the encoded audio data. Do.

  Next, when the header analysis unit 2s6 analyzes the header and finds that the data included in the input packet is video data, the controller 2s19 controls the switches SW1 and SW2 to store the video data. The packet is connected to the terminals b and d (step S211), and after the end of the header, the video encoded data area is transferred to the decode buffer 2s9 side (step S212), and the packet storing this video data is switched to the switch SW1. Is connected to the start code prefix detection unit 2s3 via the terminal b of the terminal, and in order to detect the start code of the next packet head, the activated start code prefix detection unit 2s3 is activated and the next packet start code is It is necessary to detect and separate the data of the next packet.

  At this time, however, the layer start code ('00', '00', '01', '00' to 'B8') and the packet start code ('00', ' 00 ',' 01 ', and' B9 'to' FF ') are similar to each other, which may cause confusion between the two and cause the packet boundary to be erroneous.

  Here, the operation of the present embodiment in two examples in the case where the layer start code of the video encoded data is divided between two packets will be described below using such a packet boundary as a pattern that is likely to be erroneous. This will be described with reference to FIGS. 4 and 10.

  First, as shown in FIG. 10A, 1 byte ('00') is read to the start code prefix detection unit 2s3. The start code prefix detection unit 2s3 learns from the status signal SIG8 that the current status is “0”, and the state where one “'00” is input to the start code state holding unit 2s14 by the status update signal SIG5. Notice. The start code state holding unit 2s14 updates the status from “0” to “1” in order to indicate a state where ('00') is held as history information of the head code matching state.

  Next, as shown in FIG. 10B, 1 byte ('00') is read out to the start code prefix detection unit 2s3. The start code prefix detection unit 2s3 knows that the current status is “1” by the status signal SIG8, and further notifies the start code state holding unit 2s14 by the status update signal SIG5 of the state where “00” is input. To do. The state code state holding unit 2s14 is notified of the state in which two ('00') are continuously input, and two ('00') are continuously stored as history information. In order to indicate the state, the start code state holding unit 2s14 updates the status from “1” to “2”.

  Next, as shown in FIG. 10C, 1 byte ('00') is read to the start code prefix detection unit 2s3, and the start code prefix detection unit 2s3 receives the status signal SIG8 and the current status is “2”. Knowing that there is, 1-byte signal SIG6 indicating ('00') is notified to the formatter unit 2s13. Further, the formatter unit 2s13 is activated by the status signal SIG8 indicating the status “2” from the start code state holding unit 2s14.

  Next, as shown in FIG. 10D, after that, ('01') is read to the start code prefix detection unit 2s3, and the start code prefix detection unit 2s3 is the status signal SIG8 and the current status is “2”. Knowing this, the status update signal SIG5 notifies the start code status holding unit 2s14 of the state in which ('01') is input. Thus, the start code state holding unit 2s14 is notified of the state in which ('00', '00', '01') is input, and the start code state holding unit 2s14 uses ('00 The status is updated from “2” to “3” in order to indicate the state of holding “,“ 00 ”,“ 01 ”).

  Next, as shown in FIG. 10E, ('E0') is read to the start code prefix detection unit 2s3, and the start code prefix detection unit 2s3 indicates that the current status is "3" by the status signal SIG8. Knowing that the state where ('E0') is input is notified to the start code determination unit 2s5 by the 1-byte signal SIG6. Further, the start code state holding unit 2s14 notifies the current status “3” to the start code determination unit 2s5 by the status signal SIG8.

  In response to this, the start code determination unit 2s5 determines whether it is ('00'), a value from ('00') to ('B8') based on the input 1-byte signal SIG6, or from ('B9'). The value up to ('ff') is determined from the 1-byte signal SIG6. In this example, since the input 1-byte signal SIG6 signal is ('E0'), the status is "3" and the 1-byte signal SIG6 is ('E0'). Is determined from the 1-byte signal SIG6. As shown in FIG. 11 (a), the start code determination unit 2s5 notifies the status update signal to the start code state holding unit 2s14 by SIGX1, and the status indicates that the hierarchical start code ('00') of the video encoded data is 1. The status is updated to the status “1”. On the other hand, the start code determination unit 2s5 sends a signal SIG9 to the formatter unit 2s13 according to the value of the 1-byte signal SIG6 input by the SIG9, and controls the output data of the formatter unit 2s13. The formatter unit 2s13 receives the SIG9, connects the switch SW2 to e, and prepares to transfer the formatter output to the decode buffer 2s9 via the data separation control unit 2s8. In this example, the input state of the start code prefix detector 2s3 is status “3” and the 1-byte signal SIG6 is ('E0'), that is, ('00', '00', '01', 'E0'). Since it corresponds to the pattern E in Table 1, there is no formatter output. When formatting is completed, the formatter unit 2s13 controls the controller 2s19 using the format start / end signal SIG19 so that the switch SW2 is not connected to any of the three terminals. This is because, since the start code ('00', '00', '01', 'E0') of the packet is detected by the SIG 9, the header analysis is performed after the format output process. The format output completion signal is notified to the start code determination unit 2s5 as SIGX2, and the start code determination unit 2s5 receives the SIGX2 and activates the header analysis unit 2s6 with the SIG11.

  The update of the status of the start code state holding unit 2s14 by SIGX1 is determined by which part of the start code prefix input to the decode buffer 2s9 by the formatting operation is input. When “(00)” is 1, status “1”, when two or more (“00”) are input, status “2”, (“00”, “00”, “01”) are input Status “3” is updated to status “0”.

  In the case shown in FIG. 4B, as in the case of FIG. 4A, one ('00') is output from the formatter unit 2s13 at ('00', '00'), and the status " Even when the next ('00') is input to the start code prefix detection unit 2s3, one ('00') is output from the formatter unit 2s13. Up to this point, ('00', '00'), which are part of the video layer start code at addresses γ, δ, has been transferred to the decode buffer 2s9. The subsequent operation when ('01', 'E0') is input is the same as in FIG. When ('E0') is input to the start code determination unit 2s5 by the 1-byte signal SIG6, ('00', '00') is input to the decode buffer 2s9, so that the start code state is held by SIGX1. The difference is that the status of the section 2s14 is updated from “3” to “2”.

The operation of the patterns other than the patterns A and E in Table 1 will be described below.
In the case of pattern B, after the status “1” state as shown in FIG. 10B, data (“XX”) having a value other than 1 byte (“00”) is input to the start code prefix detection unit 2s3. In this case, as shown in FIG. 11 (c), the start code prefix detection unit 2s3 determines that the input code string ('00', 'XX') is not a start code prefix. In SIG5, the start code state holding unit 2s14 is updated from the status “1” to the status “0”. The start code state holding unit 2s14 whose status is updated controls the formatter unit 2s13 to output ('00', 'XX') by the status update signal SIG8. The activated formatter unit 2s13 sends a format start signal SIG19 to the controller 2s19. The controller 2s19 connects the switch SW2 to e and outputs ('00', 'XX') to the data separation control unit 2s8. When the formatter unit 2s13 completes the output, the SIG 19 sends a format end signal to the controller unit 2s19, and switches the switch SW2 to d. The formatter end signal SIG16 is sent to the start code prefix detection unit 2s3, and the start code prefix detection unit 2s3 resumes the operation with the status “0”.

  In the case of pattern C, ('YY') having a value other than ('00') or ('01') is input to the start code prefix detection unit 2s3 in the status 2 state as shown in FIG. 10 (c). In this case, as shown in FIG. 11C, the start code prefix detection unit 2s3 determines that the input code string ('00', '00', 'YY') is not a start code prefix. Accordingly, the start code state holding unit 2s14 is updated from the status “2” to the status “0” by the status update signal SIG5. The updated start code state holding unit 2s14 controls the formatter unit 2s13 to output ('00', '00', 'YY') by the status signal SIG8. The activated formatter unit 2s13 sends a format start signal SIG19 to the controller 2s19. The controller 2s19 connects the switch SW2 to e and outputs ('00', '00', 'YY') to the data separation control unit 2s8. . When the formatter unit 2s13 completes the output, the SIG 19 sends a format end signal to the controller unit 2s19, and switches the switch SW2 to d. The formatter end signal SIG16 is sent to the start code prefix detection unit 2s3, and the start code prefix detection unit 2s3 resumes the operation with the status “0”.

  In the case of the pattern D-1, the start code prefix detection unit 2s3 is in a state other than ('00') or ('B9') to ('ff') in the status "3" as shown in FIG. When ('zz') having a value is input, as shown in FIG. 11 (c), the start code prefix detection unit 2s3 receives the input code string ('00', '00', '01', ' Since it is determined that zz ′) is not the reproduction unit start code, the start code state holding unit 2s14 is updated from the status “3” to the status “0” by the status update signal SIG5. The status-updated start code state holding unit 2s14 controls the formatter unit 2s13 to output ('00', '00', '01', 'zz') by the status signal SIG8. The activated formatter unit 2s13 sends a format start signal SIG19 to the controller 2s19. The controller 2s19 connects the switch SW2 to e, and ('00', '00', '01', 'zz') is a data separation control unit. Output to 2s8. When the formatter unit 2s13 completes the output, the SIG 19 sends a format end signal to the controller unit 2s19, and switches the switch SW2 to d. The formatter end signal SIG16 is sent to the start code prefix detection unit 2s3, and the start code prefix detection unit 2s3 resumes the operation with the status “0”.

  In the case of the pattern D-2, when ('00') is input to the start code prefix detection unit 2s3 in the status “3” state as illustrated in FIG. 10E, as illustrated in FIG. In SIG6, 1-byte data is notified to the start code discriminating unit 2s5, and the start code discriminating unit 2s5 has the input code string ('00', '00', '01', '00') as the start code of the reproduction unit. It is determined that there is a possibility, and the start code state holding unit 2s14 is updated from the status “3” to the status “4” by the status update signal SIG5. In the status “4”, the start code determination unit 2s5 further receives data other than ('00') or ('B9') to ('ff') with a 1-byte signal SIG6 from the start code prefix detection unit 2s3 (' zz ′), the status update signal SIG5 is notified to the start code state holding unit 2s14, and the status of the start code state holding unit 2s14 is updated from “4” to “0”, while the same status. The formatter unit 2s13 is activated by the update signal SIG5. The activated formatter unit 2s13 sends a format start signal SIG19 to the controller 2s19, and the controller 2s19 connects the switch SW2 to e and ('00', '00', '01', '00', 'zz'). The data is output to the data separation control unit 2s8. When the formatter unit 2s13 completes the output, the SIG 19 sends a format end signal to the controller unit 2s19 and switches the switch SW2 to d. The formatter end signal SIG16 is sent to the start code prefix detection unit 2s3, and the start code prefix detection unit 2s3 resumes the operation with the status “0”.

  In the case of the pattern D-3, when the status “4” is received and (“XY”) other than (“00”) and (“01”) is continuously notified in the SIG6, the formatter unit 2s13 The only difference is that ('00', '00', '01', '00', '00', 'XY') is output.

  Further, in the case of D-4, in the state of status “4”, after SIG6 ('00', '00', '01') and ('B9') to ('ff') are taken ('XY' ') Is continuously notified, the formatter unit 2s13 outputs (' 00 ',' 00 ',' 01 '), and the status of the start code state holding unit 2s14 is changed from status "4" to SIGX1. The only difference is that it is updated to “3”.

  As described above, according to the first embodiment, the formatter 2s13 is provided, and a code string that matches the first part ('00') of the predetermined code string detected by the start code prefix detection unit 2s3 is detected. In this case, the start code prefix detection unit 2s3 detects the remaining portion ('00', '00', '01', 'E0') of the detected predetermined code string, and ('00', '00', '00') is detected, the formatter 2s13 outputs one ('00'), and after the boundary of the packet is determined, the data not transferred to the decode buffer 2s9 Among them, data corresponding to a code string other than the code string ('00', '00', '01', 'E0') for indicating the boundary of the packet is output to the decode buffer 2s9. The input buffer is read when separating multiple encoded signals. This eliminates the need for complicated control such as the advance and retreat of the read address, which makes it easier to control the read address of the input buffer by the input buffer read control circuit, and the hardware scale is reduced accordingly. An apparatus for reproducing a multiplexed digital code string can be provided at low cost.

(Embodiment 2.)
Next, an encoded signal reproduction device according to Embodiment 2 of the present invention will be described. In the second embodiment, in addition to the functions shown in the first embodiment, it is not necessary to hold a large amount of reproduction information in the apparatus, and hardware such as a memory necessary for holding is minimized. Is made possible. Since the basic configuration is the same as that shown in FIG. 1, the description thereof is omitted here.

  FIG. 5A shows a unit structure of a packet composed of a packet header of video (video) and video encoded data following the packet header. Video encoded data included in a video packet may include a plurality of video playback basic units. In FIG. 5, video playback basic unit data 0 to 3 are included. The video playback basic unit includes a video playback basic unit start code at the head thereof. It is not guaranteed that the video playback unit start code is always included in the head portion of the video encoded data in the video packet, and the video packet can be contained in the immediately preceding video packet like the video playback basic unit data 0 in FIG. It is common that the data of the reproduction unit that has not been located is positioned at the head of the video encoded data. The display time information (PTS) is assigned to the basic unit data including the first video playback basic unit start code among the video playback units included in the video encoded data of the video packet. In the case of FIG. Basic unit data 1 corresponds to this. No display time information is given for the others. In the case of a packet having no PTS, no PTS is assigned to any video reproduction basic unit data included in the video encoded data. Of the reproduction information extracted by the header analysis unit 2s6, only information for synchronizing the reproduction time of audio and video called PTS is temporarily stored in the reproduction information holding unit 2s7. When the analysis of the header part is completed, the header analysis unit 2s6 activates the data separation control unit 2s8 and starts transferring the video data part to the decode buffer 2s9. At this time, in order to detect the boundary start code included in the video encoded data, the start code prefix detection unit 2s3 that analyzes the packet header is activated.

  When the start code prefix detection unit 2s3 detects the start code prefix code included in the video data, the start code prefix detection unit 2s3 activates the start code determination unit 2s5, and the boundary start code of the detected video data Whether the start code of the video data is the boundary start code or the start code of the playback basic unit of the video encoded data is determined by looking at the code next to the preceding code.

  If it is determined from the code following the start code prefix that the start code is the basic playback unit, the formatter 2s13 displays the display start information (PTS) that is the playback information stored in the playback information holding unit 2s7. Is added to the tail of the start code ('00', '00', '01', '00') of the reproduction basic unit based on the analysis result of the start code determination unit 2s5. That is, as shown in FIG. 5 (b), a state flag and a PTS are added to the tail of the plurality of playback basic units included in the encoded video data that are temporally located at the head, Since a flag indicating validity is assigned to the flag and PTS does not appear until the next packet header is detected in the subsequent playback basic unit, a code indicating that PTS is invalid for each of the above status flags. Attach and add PTS.

  As described above, according to the second embodiment, the formatter 2s13 is provided, and among the reproduction information included in the packet header, only the PTS that is the display time information is taken into the reproduction information holding unit 2s7, and the video is reproduced for each reproduction basic unit. Since PTS is added to the end of the start code of the playback basic unit and information (flag) indicating PTS invalidity / validity is individually added, only PTS is temporarily stored in the playback information included in the packet header. It suffices if the data is stored inside the playback device, and the data before decoding is stored in the RAM (decode buffer 2s9) outside the device, and therefore there is no need to store a large amount of playback information inside the device. Therefore, it is possible to minimize the scale of hardware such as memory required for holding playback information, and The apparatus for reproducing the barrel code sequence can be provided at low cost.

  Although it is possible to add the extracted PTS only to the first playback basic unit and not to add it to other playback basic units, the method of uniformly adding the PTS to each playback basic unit is more preferable. There is an advantage that the operation of the formatter is made constant, the control is simple, and the reproduction routine can be simplified.

(Embodiment 3)
Next, an encoded signal reproduction device according to Embodiment 3 of the present invention will be described. In the third embodiment, in addition to the functions shown in the first embodiment, the data bus is sequentially transferred in a pipeline manner by using the formatter unit to supplement the data so that the transfer control of the data bus can be simplified. It is a thing. Since the basic configuration is the same as that shown in FIG. 1, the description thereof is omitted here. Here, the start code determination unit 2s5 realizes a termination code detection means for detecting a code string indicating the termination of the encoded data.

  FIG. 6A shows a code string pattern at the end of video encoded data in multiplexed encoded data. As is clear from FIG. 6A, the end of the video encoded data is a data pattern that is less than the data width of the pipeline output from the multiplexed signal demultiplexing unit 2s or the decoding unit 2s10. The input data pattern is detected by the start code prefix detection unit 2s3 ('00', '00', '01'), and the start code determination unit 2s5 detects the boundary start code ('00') indicating the final video data. , 00 ',' 01 ',' b7 '). ('B7') indicates a sequence end code. As described above, when the start code determination unit 2s5 detects the final part of the final video data, the start code determination unit 2s5 notifies the formatter 2s13 to that effect. The formatter 2s13 generates final video data, adds padding data ('FF') to the final data code string, as shown in FIG. 6B, and transfers the data to the decode buffer 2s9. Here, when the byte width is 4 bytes (32 bits), the number of padding data to be added is about the bus width data including the video final data by inserting three or more after the final data code string. It can be pipelined. That is, assuming that the width of the data bus for pipeline transfer is n bytes, when the video data of the final part exists at the mth byte from the bus width boundary, (n−m) or more padding data is added. Thus, the video data of the last part can be read out by pipeline processing. Due to the addition of the padding data, the video final data portion that is less than the data bus width is aligned to fit within the data bus width.

  At this time, as the padding data, it is necessary to read data from the decode buffer 2s9 and select data that does not cause the decoding unit that performs decoding to malfunction. For example, there is no emulation of the start code prefix ('ff'). Then, by adding padding data to the final data string, a code string including a code that does not satisfy the pipeline data bus width necessary for decoding needs to be controlled for complicated data bus transfer such as byte access. It is possible to transfer without any problem.

  By the way, in the case of image encoded data, a code indicating the end of the program called a program end code is usually added to the end of the packet. For example, in the current DVD-Video standard, this code is added. Since it is not defined, if there is no description of this code, the host CPU will not be able to determine whether the data has been sent to the decoding buffer to the end or it has been interrupted. May not be able to be displayed and the system may freeze. In addition, when there is no sequence end code at the end of the encoded video data, there may be a problem that the last several pictures are not output.

  Such a case can be dealt with by uniformly inserting a unique code string in advance in the last packet (end of sequence) of the packet string.

  That is, FIG. 7 is a diagram showing a code string unique to the multiplexed code string inserted at the packet boundary. Here, the unique code string is assumed to have a code string similar to the packet start code. That is, this unique code string takes the form ('00', '00', '01', 'XX'). However, ('XX') is a code that is not confused with the packet start code, for example, a sequence end code ('b7') or a program end code ('b9'). The unique code string is inserted after the packet data of the last packet in the reproduction sequence by the host CPU 2s12.

  When a unique code string is input after the specific data packet by the host CPU 2s12, it is confirmed that the unique code data is input by the start code prefix detection unit 2s3 and the start code determination unit 2s5 in the multiple signal separation unit 2s. Detected. At this time, the start code determination unit 2s5 uses the notification signal (SIG20) to transfer the encoded data included in the packet before the unique code string to the decode buffer 2S9, to the data separation control unit 2s8. Make a notification.

  In response to this, the data separation control unit 2s8 performs data transfer of the packet data to the decode buffer 2s9, and when the transfer is completed, the data separation control unit 2s8 notifies the start code determination unit 2s5 of the transfer completion using the notification signal (SIG20). Upon receiving the transfer completion notification, the start code determination unit 2s5 detects that this unique code string has been detected, and that the host control signal A indicates that the data included in the packet preceding the unique code string has been transferred to the decode buffer 2s9. (SIG13) is used to notify the host CPU 2s12. Thus, the external host CPU 2s12 can reliably detect that specific packet data is stored in the decode buffer 2s9.

  Then, by identifying the inserted specific code string with respect to the code data with the specific code string inserted as described above, and adding a predetermined number of padding data after that, each bus width The size of data existing within the boundary can be made uniform. If the program end code is originally given to the last packet of the packet sequence, this will be detected in succession by adding the specific code sequence, but in actual operation, There will be no hindrance.

  As described above, according to the third embodiment, the formatter 2s13 has a data padding function, and the formatter 2s13 adds padding data to the end of the code string that does not satisfy the data bus width of the pipeline processing. Since the size of the data existing within the bus width boundary is the same, the data transfer in the pipeline in the playback device can be complicated up to the final part of the encoded data that is less than the data bus width in the pipeline. It is possible to realize the transmission without the need for complicated transfer control, and it is possible to reliably flow the data of the last part of the code data in the pipeline of the reproducing apparatus without using complicated data transfer control. .

  In addition, by adding a specific code string uniformly to the last packet of the packet sequence, when a signal of a specific data sequence is input, it is recognized as an input end of multiplexed data, and a specific code sequence is input last. After that, the data before the specific code string is input is transferred to the decode buffer, and then the detection of the specific code string is notified to an external host CPU, whereby the data of the specific packet is transferred to the decode buffer. The external CPU can reliably detect the input. Therefore, after that, the CPU 2s12 can initialize the decoding unit 2s10 by clearing the pipeline, and can prepare for the next encoding.

  In the first to third embodiments, a controller is provided in addition to the host CPU. However, as shown in FIG. 8, the controller is omitted by providing the host CPU with the function of the controller. The same effects as those of the first to third embodiments may be obtained.

  The present invention relates to an encoded signal reproducing apparatus, and in particular, simplification of the circuit configuration of an apparatus that reproduces a digital code string in which audio, video, and other additional information are multiplexed, such as video CD, DVD, and digital CS broadcasting. Useful for.

It is a block diagram which shows the structure of the coding signal decoding apparatus by Embodiment 1 thru | or 3 of this invention. It is a figure which shows the flow for demonstrating operation | movement of the coding signal decoding apparatus by the said Embodiment 1. FIG. It is explanatory drawing which shows the structure of a multiplexed signal. It is explanatory drawing which shows the multiplexed signal which is easy to make an error in the packet boundary in a decoding buffer. It is explanatory drawing which shows the multiplexed signal which is easy to make an error in the packet boundary in a decoding buffer. It is a figure which shows the relationship between a video packet and a video reproduction unit. It is a figure which shows the relationship between a video packet and a video reproduction unit. It is a figure for demonstrating the padding process of the video encoding data last part by a formatter. It is a figure for demonstrating the padding process of the video encoding data last part by a formatter. It is a figure which shows the unique code sequence inserted in the multiplex coding data. It is a block diagram which shows the modification of the coding signal decoding apparatus by Embodiment 1 thru | or 3 of this invention. It is a block diagram which shows the notional structure of the coding signal decoding apparatus of this invention. FIG. 10 is a diagram illustrating a flow for explaining operations centering on a formatter and a start code state holding unit when an input packet is a video packet in the encoded signal decoding apparatus of the present invention. FIG. 10 is a diagram illustrating a flow for explaining operations centering on a formatter and a start code state holding unit when an input packet is a video packet in the encoded signal decoding apparatus of the present invention. FIG. 10 is a diagram illustrating a flow for explaining operations centering on a formatter and a start code state holding unit when an input packet is a video packet in the encoded signal decoding apparatus of the present invention. FIG. 10 is a diagram illustrating a flow for explaining operations centering on a formatter and a start code state holding unit when an input packet is a video packet in the encoded signal decoding apparatus of the present invention. FIG. 10 is a diagram illustrating a flow for explaining operations centering on a formatter and a start code state holding unit when an input packet is a video packet in the encoded signal decoding apparatus of the present invention. FIG. 10 is a diagram illustrating a further flow for explaining operations centering on a formatter and a start code state holding unit when an input packet is a video packet in the encoded signal decoding apparatus of the present invention. FIG. 10 is a diagram illustrating a further flow for explaining operations centering on a formatter and a start code state holding unit when an input packet is a video packet in the encoded signal decoding apparatus of the present invention. FIG. 10 is a diagram illustrating a further flow for explaining operations centering on a formatter and a start code state holding unit when an input packet is a video packet in the encoded signal decoding apparatus of the present invention. FIG. 10 is a diagram illustrating a further flow for explaining operations centering on a formatter and a start code state holding unit when an input packet is a video packet in the encoded signal decoding apparatus of the present invention. It is a figure for demonstrating the read-out operation | movement of the video encoding data by the conventional encoding signal reproduction | regeneration apparatus.

Explanation of symbols

2 multiplex encoded signal reproduction device 2s multiplex signal separation unit 2s1 multiplex encoded signal supply unit 2s2 input buffer 2s3 start code prefix code detection unit 2s4 input buffer read control circuit 2s5 start code discrimination unit 2s6 header analysis unit 2s7 reproduction information holding unit 2s8 Data separation control unit 2s9 Decoding buffer 2s10 Decoding unit 2s11 Work memory 2s12 Host CPU
2s13 formatter 2s14 start code state holding unit 2s19 controller SW1 switch SW2 switch HB host bus

Claims (3)

A specific code string insertion means for inserting a specific code string into the last packet of the encoded data regardless of whether or not a code string indicating the end of the encoded data is already included;
A code string indicating the end of the encoded data from the code string of the encoded data or / and a termination code string detecting means for detecting the specific code string ;
When the code string indicating the end of encoded data or / and the specific code string is detected by the end code string detecting means, the data bus width of the pipeline transfer including the end of the encoded data is set to other A formatter unit for adding a predetermined number of pseudo data to the code string indicating the end of the encoded data or the tail of the specific code string so as to be equal to the bus width of pipeline transfer including data ;
A decode buffer for sequentially storing pipeline transferred code strings ;
An encoded signal reproducing apparatus characterized by the above. The encoded signal reproducing apparatus according to claim 1,
The input code string is a multiplex encoded signal in which audio and video, and reproduction information associated therewith are multiplexed.
Marks Goka signal reproducing apparatus you wherein a. The encoded signal reproducing apparatus according to claim 1,
The specific code string insertion means is a host CPU,
When the specific code string is detected by the terminal code string detection means and the code string to which the predetermined number of pseudo data is added is transferred to the decode buffer, the host CPU is notified.
Marks Goka signal reproducing apparatus you wherein a.

Images