SMPTE RP 2025-2007 SMPTE RECOMMENDED PRACTICE VC-1 Bitstream Storage in the ISO Base Media File Format《SMPTE推荐操作规程 ISO基媒体文件格式的VC-1位流储存》.pdf

1、 Table of Contents Page Foreword . 1 Introduction . 1 1 Scope 2 2 Conformance Notation 2 3 Normative References 2 4 VC-1 Track definition. 2 5 VC-1 Sample definition . 3 5.1 Random access points 3 6 VC1SampleEntry definition . 3 7 VC1SpecificBox definition. 5 8 VC1DecSpecStruc definition. 5 8.1 Over

2、all definition . 5 8.2 VC1SequenceHeader_B definition . 6 8.3 VC1SequenceHeader_C definition. 6 8.4 VC1AdvDecSpecStruc definition 8 Foreword SMPTE (the Society of Motion Picture and Television Engineers) is an internationally-recognized standards developing organization. Headquartered and incorporat

3、ed in the United States of America, SMPTE has members in over 80 countries on six continents. SMPTEs Engineering Documents, including Standards, Recommended Practices and Engineering Guidelines, are prepared by SMPTEs Technology Committees. Participation in these Committees is open to all with a bon

4、a fide interest in their work. SMPTE cooperates closely with other standards-developing organizations, including ISO, IEC and ITU. SMPTE Engineering Documents are drafted in accordance with the rules given in Part XIII of its Administrative Practices. This SMPTE Recommended Practice was prepared by

5、Technology Committee C24. Introduction This section is entirely informative and does not form an integral part of this Recommended Practice. This document was initiated by C24 in response to the need to store VC-1 coded bitstreams in file formats derived from the ISO Base Media File Format such as t

6、he MP4 file format and the 3GPP file format. Page 1 of 9 pages RP 2025-2007 SMPTE RECOMMENDED PRACTICE VC-1 Bitstream Storage in the ISO Base Media File Format Copyright 2007 by THE SOCIETY OF MOTION PICTURE AND TELEVISION ENGINEERS 595 W. Hartsdale Ave., White Plains, NY 10607 (914) 761-1100 Approv

7、ed April 4, 2007 RP 2025-2007 Page 2 of 9 pages 1 Scope The purpose of this specification is to define the necessary structures for the integration of VC-1 coded bitstreams in a file format that is compliant with the ISO Base Media File Format. Examples of file formats that are derived from the ISO

8、Base Media File Format include the MP4 file format and the 3GPP file format. 2 Conformance Notation Normative text is text that describes elements of the design that are indispensable or contains the conformance language keywords: “shall“, “should“, or “may“. Informative text is text that is potenti

9、ally helpful to the user, but not indispensable, and can be removed, changed, or added editorially without affecting interoperability. Informative text does not contain any conformance keywords. All text in this document is, by default, normative, except: the Introduction, any section explicitly lab

10、eled as “Informative“ or individual paragraphs that start with “Note:” The keywords “shall“ and “shall not“ indicate requirements strictly to be followed in order to conform to the document and from which no deviation is permitted. The keywords, “should“ and “should not“ indicate that, among several

11、 possibilities, one is recommended as particularly suitable, without mentioning or excluding others; or that a certain course of action is preferred but not necessarily required; or that (in the negative form) a certain possibility or course of action is deprecated but not prohibited. The keywords “

12、may“ and “need not“ indicate courses of action permissible within the limits of the document. The keyword “reserved” indicates a provision that is not defined at this time, shall not be used, and may be defined in the future. The keyword “forbidden” indicates “reserved” and in addition indicates tha

13、t the provision will never be defined in the future. 3 Normative References The following standards contain provisions which, through reference in this text, constitute provisions of this recommended practice. At the time of publication, the editions indicated were valid. All standards are subject t

14、o revision, and parties to agreements based on this recommended practice are encouraged to investigate the possibility of applying the most recent edition of the standards indicated below. SMPTE 421M-2006, Television VC-1 Compressed Video Bitstream Format and Decoding Process ISO/IEC 14496-12:2005,

15、Information Technology Coding of Audio-Visual Objects Part 12: ISO Base Media File Format 4 VC-1 Track definition In the terminology of the ISO Base Media File Format specification (ISO/IEC 14496-12), a VC-1 track is a video track. It therefore follows that these rules apply to the media box in the

16、VC-1 track: In the Handler Reference box, the handler_type field shall be set to vide. The Media Information Header box shall contain a Video Media Header box. The Sample Description Box shall contain a box derived from VisualSampleEntry. (For VC-1 tracks, this box is called VC1SampleEntry and is de

17、fined in section 6.) RP 2025-2007 Page 3 of 9 pages As is the case with all video tracks in ISO Base Media File Format, references to VC-1 samples in the Sample Table Box shall be arranged in Decoding Time order. Decoding Time order is equivalent to the concept of “coded order” in SMPTE 421M. 5 VC-1

18、 Sample definition A VC-1 sample is defined as follows: For Simple and Main profiles: Exactly one VC-1 frame. For Advanced profile: One or more Encapsulated Bit-stream Data Units (EBDUs), as defined in Annex E of SMPTE 421M. The VC-1 sample shall contain data belonging to exactly one VC-1 frame. In

19、the case of interlaced video, a VC-1 frame consists of two fields that may be coded as separate pictures. The two pictures still belong to the same VC-1 frame and hence also to the same VC-1 sample. It shall be noted that when Advanced profile is used, a VC-1 sample may contain EBDUs that are not pa

20、rt of a VC-1 frame, such as a Sequence Header EBDU. The following rules apply to such EBDUs: If a Sequence Header is present in a VC-1 sample, it shall be the first EBDU in the sample. If an Entry-point Header is present in a VC-1 sample, it shall be the first EBDU in the sample, unless a Sequence H

21、eader is also present, in which case the Entry-point Header shall follow directly after the end of the Sequence Header and any Sequence-level User Data. The EBDU Start Code field is used to identify the type of the EBDU. The list of values for the EBDU Start Code field is given in SMPTE 421M, Annex

22、E.5. VC-1 samples shall be byte-aligned. If necessary, up to 7 zero-valued padding bits shall be added to the end of a VC-1 sample to achieve byte-alignment. The Padding Bits Box (defined in section 8.23 of ISO/IEC 14496-12) need not be used to record padding bits that are added to a sample to align

23、 its size to the nearest byte boundary. 5.1 Random access points For Simple and Main profiles, each VC-1 sample that contains an I-picture shall be a random access point. For Advanced profile, a VC-1 sample shall be a random access point if any of the following conditions are met: The sample contain

24、s both a Sequence Header and an Entry-point Header. The sample contains an Entry-point Header and the no_multiple_seq bit in VC1AdvDecSpecStruc is 1. The definition of VC1AdvDecSpecStruc is given in section 8.4. The Sync Sample Box shall be used to mark all random access points in the track. The Syn

25、c Sample Box is defined in section 8.20 of ISO/IEC 14496-12. 6 VC1SampleEntry definition The box type of the VC1SampleEntry Box shall be vc-1. The VC1SampleEntry Box is defined as follows: RP 2025-2007 Page 4 of 9 pages VC1SampleEntry := BoxHeader Reserved_6Data-reference-index Reserved_16 Width Hei

26、ght Reserved_4 Reserved_4Reserved_4 Reserved_2Reserved_32 Reserved_2 Reserved_2VC1SpecificBox Table 1 VC1SampleEntry fields Field Type Details Value BoxHeader.Size Unsigned int(32) BoxHeader.Type Unsigned int(32) vc-1 Reserved_6 Unsigned int(8) 6 0 Data-reference-index Unsigned int(16) Index to a da

27、ta reference that to use to retrieve the sample data. Data references are stored in data reference boxes. Reserved_16 Const unsigned int(32) 4 0 Width Unsigned int(16) Maximum visual width of the stream, in pixels Height Unsigned int(16) Maximum visual height of the stream, in pixels Reserved_4 Cons

28、t unsigned int(32) 0x00480000 Reserved_4 Const unsigned int(32) 0x0048000Reserved_4 Const unsigned int(32) 0 Reserved_2 Const unsigned int(16) 1 Reserved_32 Const unsigned int(8) 32 0 Reserved_2 Const unsigned int(16) 24 Reserved_2 Const int(16) -1 VC1SpecificBox Information specific to the VC-1 dec

29、oder. The layout of the VC1SampleEntry box is identical to that of VisualSampleEntry defined in ISO/IEC 14496-12 (including the reserved fields and their values), except that VC1SampleEntry ends with a box containing VC-1 decoder initialization parameters called VC1SpecificBox. The VC1SpecificBox fi

30、eld structure for VC-1 is defined in section 7. Width and Height: specify the maximum horizontal and vertical sizes, respectively, of the displayed frames. RP 2025-2007 Page 5 of 9 pages 7 VC1SpecificBox definition VC1SpecificBox shall always be included in the VC1SampleEntry box. VC1SpecificBox is

31、composed of the following fields: VC1SpecificBox := BoxHeader DecSpecificInfo Table 2 The VC1SpecificBox fields Field Type Details Value BoxHeader.Size Unsigned int(32) BoxHeader.Type Unsigned int(32) dvc1 DecSpecificInfo VC1DecSpecStruc Structure which holds the VC-1 decoder initialization paramete

32、rs BoxHeader Size and Type: specify the size and type of the VC1SpecificBox. The type must be dvc1. DecSpecificInfo: A structure of the VC1DecSpecStruc type which is used to store the VC-1 decoder initialization parameters. VC1DecSpecStruc is defined in section 8. 8 VC1DecSpecStruc definition 8.1 Ov

33、erall definition VC1DecSpecStruc is defined as follows: struct VC1DecSpecStruc unsigned int (4) profile; unsigned int (3) level; const bit (1) reserved = 0; if (profile = 0 | profile = 4) struct VC1SequenceHeader_C struct_C; struct VC1SequenceHeader_B struct_B; else if (profile = 12) struct VC1AdvDe

34、cSpecStruc struct_AP; The definitions of VC1DecSpecStruc members are as follows: profile: specifies the VC-1 encoding profile used to produce the coded bitstream. It shall be 0 for VC-1 Simple profile, 4 for VC-1 Main profile, and 12 for VC-1 Advanced profile. All other values are SMPTE reserved (as

35、 defined in SMPTE 421M). level: the highest encoding level used in the VC-1 bit stream. For VC-1 Simple profile, level shall be 0 to indicate Low level and 2 to indicate Medium level. For VC-1 Main profile, level shall be 0 to indicate Low level, 2 to indicate Medium level and 4 to indicate High lev

36、el. For VC-1 Advanced profile, level shall take a value from 0 through 4, corresponding to VC-1 Advanced profile levels L0 through L4, respectively. All other values are SMPTE reserved (as defined in SMPTE 421M). RP 2025-2007 Page 6 of 9 pages struct_C: A structure of the VC1SequenceHeader_C type. T

37、his structure is identical to STRUCT_SEQUENCE_HEADER_C as defined in Annex J of SMPTE 421M. struct_B: A structure of the VC1SequenceHeader_B type. This structure is identical to STRUCT_SEQUENCE_HEADER_B as defined in Annex J of SMPTE 421M. struct_AP: A structure of the VC1AdvDecSpecStruc type. It is

38、 defined in section 8.4. 8.2 VC1SequenceHeader_B definition VC1SequenceHeader_B is defined as follows: struct VC1SequenceHeader_B unsigned int (3) level; unsigned bit (1) cbr; const bit (4) res1 = 0; unsigned int (24) hrd_buffer; unsigned int (32) hrd_rate; unsigned int (32) framerate; The definitio

39、ns of VC1SequenceHeader_B members are as follows: level: shall be set to the same value as is used for the level field in VC1DecSpecStruc (see section 8.1). cbr: shall be 1 if the content was generated using a constant bit rate model. If it was generated using any other bit rate model, cbr shall be

40、0. res1: is reserved and shall be set to zero. hrd_buffer: specifies the buffer size B of the hypothetical reference decoder (HRD) in millisecond units. (The HRD is defined in Annex C of SMPTE 421M.) hrd_rate: specifies the peak transmission rate R of the hypothetical referenced decoder (HRD) in bit

41、s per second. framerate: shall be set to the rounded frame rate (fps) of the track. The framerate field shall be set to 0xffffffff if the frame rate is not known, unspecified, or non-constant. 8.3 VC1SequenceHeader_C definition VC1SequenceHeader_C is defined as follows: struct VC1SequenceHeader_C un

42、signed int (4) profile; unsigned int (3) frmrtq_postproc; unsigned int (5) bitrtq_postproc; unsigned bit (1) loopfilter; const bit (1) reserved1 = 0; unsigned bit (1) multires; const bit (1) reserved2 = 1; unsigned bit (1) fastuvmc; unsigned bit (1) extended_mv; unsigned int (2) dquant; unsigned bit

43、 (1) vstransform; const bit (1) reserved3 = 0; RP 2025-2007 Page 7 of 9 pages unsigned bit (1) overlap; unsigned bit (1) syncmarker; unsigned bit (1) rangered; unsigned int (3) maxbframes; unsigned int (2) quantizer; unsigned bit (1) finterpflag; const bit (1) reserved4 = 1; The definitions of VC1Se

44、quenceHeader_C members are as follows: profile: shall be set to the same value as is used for the profile field in VC1DecSpecStruc (see section 8.1). frmrtq_postproc: signals quantized frame rate information as defined in Annex J.1.7 of SMPTE 421M. bitrtq_postproc: signals quantized frame rate infor

45、mation as defined in Annex J.1.8 of SMPTE 421M. loopfilter: shall be set to 1 if loop filtering is enabled and Main profile is used. If loop filtering is not enabled, the value shall be 0. The value shall be 0 when Simple profile is used. reserved1: shall be set to 0 and other values shall be forbid

46、den. multires: shall be set to 1 if the coded frames may be coded at a lower resolution than specified by the width and height fields in VC1SampleEntry, otherwise it shall be set to 0. See Annex J.1.10 of SMPTE 421M for additional constraints on the use of this field. reserved2: shall be set to 1 an

47、d other values shall be forbidden. fastuvmc: controls the subpixel interpolation and rounding of color-difference motion vectors as defined in Annex J.1.11 of SMPTE 421M. The fastuvmc field shall be 1 when Simple profile is used. extended_mv: shall be set to 1 if extended motion vectors may be prese

48、nt in P- and B-pictures. The extended_mv field shall be 0 when Simple profile is used. dquant: indicates if the quantization step size may vary within a frame, as defined in Annex J.1.13 of SMPTE 421M. vstransform: shall be set to 1 if variable-sized transform coding is enabled. It shall be set to 0

49、 if variable-sized transform coding is not enabled. reserved3: shall be set to 0 and other values shall be forbidden. overlap: shall be set to 1 if overlapped transforms may be used. It shall be set to 0 if overlapped transforms are not used. syncmarker: shall be set to 1 if synchronization markers may be present in the bit stream and Main profile is used. The value shall be 0 if synchronization markers are not present in the bit stream. The syncmarker field shall be 0 when Simple profile is used.