SMPTE ST 353M-2000 Transport of MEPG-2 Recording information as Ancillary Data Packets.pdf

《SMPTE ST 353M-2000 Transport of MEPG-2 Recording information as Ancillary Data Packets.pdf》由会员分享，可在线阅读，更多相关《SMPTE ST 353M-2000 Transport of MEPG-2 Recording information as Ancillary Data Packets.pdf（10页珍藏版）》请在麦多课文档分享上搜索。

1、SMPTE STANDARD for Television - Transport of MPEG-2 Recoding Information as Ancillary Data Packets SMPTE 353M-2000 Contents 1 Scope 2 Normative references 3 General 4 Definitions 5 Ancillary data packet formatting for transport of Annex A Using the MPEG-2 video recoding data set Annex B Bibliography

2、 MPEG-2 video recoding data set 1 Scope This standard specifies the mechanism for the transport of MPEG-2 video recoding information as ancillary data packets in an ancillary data space - for example, through ITU-R BT.656 / SMPTE 259M interfaces. The video recoding information transported through th

3、is mechanism is for the purpose of preserving picture quality at re-encoding stages when cascading MPEG-i! decoders and encoders. Although the speciied mechanism operates on 1 O-bit digital video intedaces, it is by design transparent to systems limited to 8-bit operation. 2 Normative references The

4、 following standards contain provisions which, through reference in this text, constitute provisions of this standard. At the time of publication, the editions indicated were valid. All standards are subject to revision, and parties to agreements based on this standard are encouraged to investigate

5、the possibility of applying the most recent edition of the standards indicated below. Page 1 of 10 pages SMPTE 291M-1998, Television - Ancillary Data Packet and Space Formatting SMPTE 327M-2000, Television - MPEG-2 Video Recoding Data Set SMPTE 329M-2000, Television - MPEG-2 Video Recoding Data Set

6、- Compressed Stream Format 3 General The principal application of this standard is to allow the preservation of picture quality when cascading video decoders and encoders following ISO/IEC 1381 8-2. This standard defines the transport of MPEG-2 video recoding information in the ancillary data space

7、pro- vided by digital video interfaces. In addition, the trans- port mechanism specified in this standard has been designed so that it can work with digital video systems in which operation is limited to 8-bit resolution. The following shall apply: - The transported MPEG-2 video recoding infor- mati

8、on shall be compliant with the MPEG-2 video recoding data set as defined in SMPTE 327M. - This data set shall be formatted according to the stream format defined in SMPTE 329M. - This formatted data set shall be transported in the form of ancillary (ANC) data packets as speci- fied in SMPTE 291M. Th

9、e transport mechanism specified in this standard shall be compliant with SMPTE 291 M. Part of both the Copyright 02000 by THE SOCI OF MOTION PICTURE AND TELEVISION ENGINEERS 595 W. Haitsdale Ave., White Plains, NY 10607 (914) 761-1100 Approved October 10,2000 vertical blanking ancillary data space,

10、V-ANC, and the horizontal blanking ancillary data space, H-ANC, shall be used. The V-ANC space shall be used to carry picture rate information only. This is the most basic, highest prior- ity, part of the recoding data set. For low bit rate, long GOP applications in particular, it typically brings t

11、he largest picture quality improvement step while further refinements are brought when more information is available. The H-ANC space shall be used to carry the other part of the recoding data set. The use of the reduced bandwidth indicator, as specified in SMPTE 329M, will allow the transmission, m

12、ore or less, of this part of the recoding data set, depending upon the transmission capacity available in the H-ANC space. 4 Definitions 4.1 aligned(temporally): In the scope of this standard, it means temporally aligned; arranged so that information related to a picture (frame or field) is inserted

13、 within the blanking interval of that decoded picture. 4.2 cascading: The process where video that has once been coded (compressed) is sub- sequently decoded and coded once more. This cascaded step could carry on for any number of generat ions. 4.3 coding: The process by which an uncorn- pressed vid

14、eo sequence is compressed to a bit stream that conforms to ISOAEC 13818-2. 4.4 decoder: A compressed bitstream decoder that complies with ISOAEC 13818-2. 4.5 MPEG profile/level: As defined in ISOAEC 1381 8-2. 4.6 MPEG-2 video recoding data stream: A b it- stream conveying the MPEG-2 video recoding d

15、ata set in the compressed stream format as defined in SMPTE 329M. 4.7 picture: As defined in ISOAEC 13818-2. 4.8 picture rate information: In this standard, this corresponds to the coding information from the ISO/IEC 1381 8-2 bitstream that relates to the 2000 8357LiOL OOOLi287 bLO m whole picture,

16、plus some additional information derived from sequence and GOP headers, but excluding slice() and macroblock() information. 4.9 recodingdataaccessunit: A section of the MPEG-2 video recoding data stream which con- tains in compressed format the recoding data of a single picture including any sequenc

17、e and GOP headers that could precede that picture. 4.10 recodingdataset: As defined in SMPTE 327M. 4.11 transcoding: A conversion within the MPEG-2 video elementary stream domain, such as bit-rate changing or changing the group of pictures (GOP) structure. 4.12 video: In this standard, a signal conf

18、orming to ITU-R BT.601. 4.13 video access unit: As defined in ISO/IEC 1381 8-1. 4.14 video elementary stream: An ordered series of bits conforming to ISOIIEC 1381 8-2. 5 Ancillary data packet formatting for trans- port of MPEG-2 video recoding data set 5.1 Ancillary video recoding data packet struc-

19、 ture An ancillaiy video recoding data packet shall be of constant length. It shall contain MPEG-2 video recod- ing data preformatted as defined in SMPTE 329M. The ancillary video recoding data packets shall be type 2 and shall be transmitted in the ancillary data space as defined in SMPTE 291 M (se

20、e figure 1). 5.2 Ancillary video recoding data packet format- ting The ADE DID, SDID, DC, and CS words shall be compliant with SMPTE 291 M. The recoding data control (RDC), recoding data block number (RDBN), recoding data block extension (RDBE), recoding data bandwidth information (RDBI), and cyclic

21、 redundancy check (CRC) words are further defined below (see table 1). Page 2 of 10 pages STD-SMPTE 353M-ENGL 2000 W 8357401 000288 557 W SMPTE 353M-2000 Ancillary recoding data packet () =- -LLLL m m Recoding Data Payload OLL on (I 2011 24 words) oc3m I I , , ADF Number of words DID DC RDBN RDB AD

22、F ADF DID SDID DC RDC RDBN RDBE RDBI CRC cs 3 AnciUary Data Flag Data ID Secondary Data ID Data Count Re-coding Data Control Re-coding Data Block Number Ming Data Block Extension Re-coding Data Bandwidth Information Cydic Redundancy Check Checksum DC Recoding data control If (recoding-data-stariflag

23、 = 1) Recoding data block number Recodina data block extension Figure 1 - Structure of the ancillary recoding data packets Table 1 - Ancillary recoding data packet syntax 1 1 1 1 I CRC Recoding data bandwidth information 1 Recoding data payload() CRCO 2 120 / 124 ) 3 VIU I 1 SDID 1 Checksum I 1 cs 1

24、) 120 if recoding-datastartflag is one; 124 otherwise. Page 3 of 10 pages STD-SMPTE 353M-ENGL 2000 8357Y03 OOOq289 q93 SMPTE 3531111-2000 - Ancillary data flag (ADF) The ADF shall have the value OOOh 3FFh 3FFh. - Data ID (DID) The DID shall have the value of 08h for 87 through BO. B8 is even par for

25、 B7 through BO; B9 is the comple- ment of B8. - Secondary data ID (SDID) For recoding data transported into the V-ANC space, the SDID shall have the value of 08h for B7 through BO. For recoding data transported into the H-ANC space, the SDID shall have the value of OCh for 87 through BO. 88 is even

26、parity for 87 through BO; B9 is the comple- ment of B8. - Data count (DC) The DC shall have the value 80h for B7 through BO, representing the fixed 128-word recoding data packet size. B8 is even parity for B7 through BO; B9 is the comple- ment of 88. - Recoding data control (RDC) The RDC shall speci

27、fy the start and the continuity of a recoding data access unit in the recoding data stream (see table 2). The recoding-data-start-flag, when set to one, shall indicate that the recoding data payload of this recod- ing data packet will begin with the first byte of a recoding data access unit. The rec

28、od- ing-data-start-flag shall be otherwise set to zero. The V-ANC-stop-flag, when set to one, shall indicate the last ancillary recoding data packet located in the V-ANC data space. This means that the next ancillary recoding data packet (if any) for the current recoding data access unit is to be fo

29、und in the H-ANC data space. The continuity-counter is an unsigned integer value incrementing by one at each consecutive recoding data packet, starting with the value 1 for the first recoding data packet of a recoding data access unit (recoding-data-start-flag = 1). The continu- ity-counter wraps ar

30、ound to zero after its maximum value of 15, and continues again. - Recoding data block number (RDBN) The RDBN defines the number of recoding data pack- ets that are used to transport a recoding data access unit. This word shall be present in the recoding data packet only if the recoding-data-statt-f

31、lag is set to one in this packet (see table 3). For large size recoding data access units where the recoding data block number exceeds 63, an extension field is signaled by 6 extra bits. These bits are located in the RDBE word. A zero value for the recod- ing-data-block-number (including the 6 exten

32、sion bits) means unknown size. - Recoding data block extension (RDBE) The RDBE contains the 6 extension bits (MSBs) of the recoding-data-block-number. This word shall be pre- sent in the recoding data packet only if the recod- ing-data-startflag is set to one in this packet (see table 4). The value

33、of this B3 and B4 shall reproduce the rd-bw-in- dicator value present in the current recoding data access unit; Bits B5 through B7 are reserved for future use. By default, they shall be set to zero. RDBI-2, the second RDBI word, is resewed for future use. By default, this word shall be set to the va

34、lue 200h. Page 4 of 10 pages SMPTE 353M-2000 RDC (10-bit word) 89 88 87 B6 85 84 63 82 B1 BO Assign ment Not BB Even parity for data contained in RDC bit 7 through bit O Recoding-data-startflag V-AN C-cto pf lag Continuity_counter3 Co nt in u it y-c o u n t e r 21 Continuity-counterl Continu it y-c

35、o u n t e r O Set to O Set to O Table 3 - Recoding data block number 87 B6 B5 84 83 82 B1 BO RDBN (10-bit word) I Assignment B9 I Not B8 R ec o d i n g-d at a-b I oc k-n u m be r 51 Recoding-data-block-number4 Recoding-dat a-bloc k-nu rnber3 R ecoding-dat a-bloc k-n u rn ber2 Recoding-dat a-bloc k-n

36、 u m ber 1 R ec o d in g-data-b I oc k-n u mb e r( O Set to O Set to O 7 88 1 Even aaritv for data contained in RDBN bit 7 throuah bit O I RDBE (10-bit word) B9 B8 87 B6 B5 B4 83 B2 B1 BO Asci g n ment Not 88 Even parity for data contained in RDBE bit 7 through bit O Recoding-data-block-numberll R e

37、co di n g-data-b I o c k-n u mb e ri1 O Recoding-data-block-n umber9 Recoding-data-block-number8 Recoding-data-block-number7 R ec o d in g-dat a-bl oc k-n u m be r 61 Set to O Set to O Table 4 - Recoding data block extension Page 5 Of 10 pages Table 5 - RDBI-1 - Recoding data bandwidth information -

38、 First word RDBI-1 (10-bit word) B9 68 87 B6 B5 Assign ment Not 88 Even parity for data contained in RDBI bit 7 through bit O Reserved (default value = O) Reserved (default value = O) Reserved (default value = O) 84 B3 82 1 B1 T Set t O Rd-bw-indicator 11 Rd-bw-indicatorIO Rd bw flaa I BO l Set to 0

39、 I - Recoding data payload formatting The 120 or 124 recoding data payload words (RDW) of an ancillary recoding data packet are formatted as shown in tables 6 and 7, depending on the ancillary data space location. Table 6 indicates the payload word structure for recoding data packets located in the

40、V-ANC data space for which the recoding data stream shall be partitioned in 4-bit words where the MSB is the first occurring bit in the recoding data stream. The recod- ing data stream is always byte aligned (see note 1). Each byte shall be partitioned into two 4-bit recoding data words. Bits 82 and

41、 B3 of the recoding data words are reserved for future use. By default, they shall be set to zero. Table 7 indicates the payload word structure for recoding data packets located in the H-ANC data space for which the recoding data stream shall be partitioned in -bh words where the MSB is the first oc

42、curring bit in the recoding data stream. The recod- ing data stream is always byte aligned (see note 1). Each group of 3 consecutive bytes shall be partitioned into 4 consecutive words (6-bit recoding data words). The value of unused bits in each group of 3 consecu- tive bytes shall be set to zero.

43、The recoding data payload shall not contain informa- tion from more than one recoding data access unit. For the case where a recoding data access unit does not fit an entire number of recoding data packets, stuffing words set to the value 2001, shall be used to fill the last recoding data packet. Wh

44、en the recoding-data-start-flag is set to one, the recoding data packet payload shall start with the first byte of the recoding data access unit. In order to prevent overload of the ancillary data space, any stuffing byte from the recoding data stream shall be removed. NOTES 1 Recoding data are form

45、atted as a compressed stream format and are subsequently byte aligned on MPEGP start codes. 2 Forthe case where the rd-bw-indicator is less than 3, or rd-bw-flag=O, the last recoding data packet located within the V-ANC data space (V-ANC-stop-flag = 1) shall only contain picture rate information. St

46、uffing words set to the value 200h shall be used to fill this last recoding data packet of the V-ANC data space. - Cyclic redundancy check (CRC) Recoding data packets are protected with a 1-bit CRC. The CRC shall be calculated over the recoding data packet excluding ADF, DID, SDID, and DC words. The

47、 CRC applies to the entire l0-bit word (LSB first) where bits B1 and 82 shall always be set to zero (O). The CRC polynomial is: CRC(16-bit) = XI6 + XI2 + X5 + 1. The generator polynomial circuitry is described in figure 2. All registers shall be set to one (1) at initialization; .e., at the start of

48、 every packet (see table 8). Page 6 of 1 O pages - STD-SMPTE 353M-ENGL 2000 8357403 0004272 T88 RDW (1 O-bit word) B9 SMPTE 353M-2ooo Assignment Not 88 88 87 B6 Even parity for data contained in RDW bit 7 through bit 4-bit recoding data word MSB 4-bit recoding data word B5 84 83 4-bit recoding data

49、word 4-bit recoding data word LSB Reserved (default value = O) I BO I Set to O I B2 B1 Table 7 - Payload words in the recoding data packets located in H-ANC data space Reserved (default value = O) Set to O RDW (10-bit word) I Assignment B9 B8 Not B8 Even parity for data contained in RDW bit 7 through bit O 87 I 6-bit recoding data word MSB B6 85 6-bit recoding data word 6-bit recoding data word 84 83 B2- I 6-bit recoding data word LSB 6-bit recoding data word 6-bit recodino data word Bi BO Table 8 - CRC words of