1、SMPTE RECOMMENDED PRACTICE RP 213-2001 MPEG-2 Operating Ranges Table of contents 1 Scope 2 Normative references 3 Application 4 MPEG-2 video parameters 5 Audio interchange Annex A Variable bit rate and constant bit rate Annex B Interoperability Annex C Overview and bibliography of related documents
2、1 Scope This practice specifies the structure and parameters of the data for interfacing MPEG-2 4:2:2 profile and digital audio in the professional environment. The purpose of this practice is to facilitate video and audio bitstream interchange between MPEGQ compliant equipment. An overview of assoc
3、iated documents in given in annex C. The combination of this document and those refer- enced in annex C will assist design and application of M P EG -2-based prof essional television equipment that facilitates bitstream interchange among different applications and over a wide set of user requirement
4、s. This practice is limited to the video and audio parameters of such a system. This practice defines the MPEG-2 operating ranges. These ranges constrain characteristics of the MPEG-2 4:2:2 profile to ensure bitstream interchange in the professional environment. These operating ranges are subsets of
5、 ISO/MPEG profiles and levels. This practice defines two operating ranges for standard- definition television and three operating ranges for Page 1 of 8 pages highdefinition television. It also defines a hierarchial relationship among the ranges. All the MPEG-2 data structures defined in this practi
6、ce are ISOAEC 13818-2 4:2:2 profile compliant and as such are decodable by MPEG-2 4:2:2 profile compli- ant stand-alone decoders at the appropriate level. Inasmuch as the 4:2:2 profile also requires stand- alone decoders to decode main profile structures (4:2:0), existing main profile sources can be
7、 accom- modated. 2 Normative references The following standards contain provisions, which, through reference in this text, constitute provisions of this practice. At the time of publication, the editions indicated were valid. All standards are subject to revision, and parties to agreements based on
8、this practice are encouraged to investigate the possibility of applying the most recent edition of the standards indicated below. AES3-1992, Digital Audio Engineering - Serial Transmission Format for Two-Channel Linearly Represented Digital Audio Data ANSVSMPTE 293M-1996, Television - 720 x 483 Acti
9、ve Line at 59.94-Hz Progressive Scan Production - Digital Representation SMPTE 274M-1998, Television - 1920 x 1080 Scan- ning and Analog and Parallel Digital Interfaces for Multiple Picture Rates SMPTE 296M-2001, Television - 1280 x 720 Pro- gressive Image Sample Structure - Analog and Digital Repre
10、sentation and Analog Interface SMPTE 302M-1998, Television - Mapping of AES3 Data into MPEG-2 Transport Stream Copyright 2001 by THE SOCIETY OF MOTION PICTURE AND TELEVISION ENGINEERS 595 W. Hamdale Ave., White Plains, NY 10607 (914) 761-1 100 Approved October 15,2001 RP 213-2001 SMPTE 308M-1998, Te
11、levision - MPEG-2 4:2:2 Profile at High Level SMPTE 327M-2000, Television - MPEG-2 Video Recoding Data Set tiodemission bandwidth. defining an MPEG-2-based system. Specific operat- ing parameter choices will depend on the individual application requirements, including editing capability, storage cap
12、acity, contribution feeds, and distnbu- SMPTE 328M-2000, Television - MPEG-2 Video Elementary Stream Editing Information SMPTE RP 202-2000, Video Alignment for MPEG-2 Coding d ISOAEC 1381 8-2:2000, Information Technology - Generic Coding of Moving Pictures and Associated Audio Information: Video ITU
13、-R BT.601-5 (10/95), Studio Encoding Parameters of Digital Television for Standard 4:3 and Wide-Screen 16:9 Aspect Ratios ITU-R BT.709-4 (09/00), Parameter Values for the HDTV Standards for Production and Intemational Programme Exchange In making this selection for a given application environ- ment,
14、 it is further recognized that tradeoffs among many different parameters must be considered. Such considerations include the bitstream overhead im- posed by various operating range constraints, the required degree of bitstream interoperability among various types of broadcast equipment, and overall
15、system complexity. Additional information about system interoperability is given in annex B. For audio, no single worldwide compressed standard has been adopted; various transmission systems are in use depending upon geographic area. Global audio interchange can, therefore, only be achieved by speci
16、- fying a noncompressed audio format. ITU-R BT.1358 (02/98), Studio Parameters of 625 and 525 Line Progressive Scan Television Systems 4 MPEG-2 video parameters 4.1 Operating ranges 3 Application The flexibility of MPEG-2 compression allows MPEG-2- based equipment to meet the diverse operational req
17、uirements of a broad range of professional tele- vision applications. Although some applications might be served by choosing a specific operating point, differ- ent users have different constraints and objectives, and may choose different specific operating parameters. Within professional applicatio
18、ns of MPEG-2, including the HDN extensions to MPEG-2 as defined by SMPTE 308M, five operating ranges are defined in this clause as shown in figure 1. Separate long- and short-GoP ranges are defined for both main level and high level systems. Additional operating ranges may be added as required to me
19、et future HDTV require- ments. Cognizant of these considerations, this practice specifies the following: Operating ranges 1 and 2 cover the MPEG-2 4:2:2P Q ML options including the standard 525-line and 625-line SDTV formats. Operating ranges 3 and 4 cover the MPEG-2 4:2:20 HL including: - Operating
20、 ranges including constrained bit rates and group-of-pictures (GoP) structures; - Operating ranges created for random access and editing capability; 480-line progressive scan 576-line progressive scan - Spatial alignment of coded images; 720-line progressive scan 1080-line interlaced scan - Use of 4
21、8-kHz sampled digital audio. 1080-line progressive scan (up to 30-Hz frame rate) Net bit rates specified for each operating range are maximum video bit rates and shall include all datathat are included in video buffer verifier (VBV) calculation. This practice, in conjunction with documents specified
22、 in annex C, describe parameter choices available in MPEG-2 and the factors to be taken into account when Page 2 of 8 pages RP 213-2001 Up to 50 Mbtls I-oniy coding Bit-rate Up to 50 Mbs Any GOP structure Up to 300 Mbtls Operating Range 4 Operating Range 38 Up to 175 Mbtls Operating Range 3A Up to 8
23、0 Mbtls I. Operating Range 2 I SDN Operating Range 1 I SDTV Figure 1 - SMPTE operating ranges 4.1.1 Operating range 1 (SDTV - Any GoP) Operating range 1 covers SDTV coded at up to 50 MbWs and may use temporal predictive coding. For GoP structures greater than one, future equip- ment designs should c
24、onform to this operating range. This will facilitate bitstream interchange over the full range of MPEG-2 main level compliant equipment. (See note below.) than 1,668,328 bits net data or, at 25 I-frames per second, no frame may have more than 2,000,000 bits net data. For SD I-frame-only applications
25、, future equipment designs should conform to this operating range. When coupled with the bitstream constraints defined in 4.4, this will facilitate bitstream interchange over the full range of MPEG-2 main level compliant equipment. (See note below.) 4.1.2 Operating range 2 (SDTV - I-only) Operating
26、range 2 covers SDN coded at up to 50 Mbit/s, using no temporal predictive coding. For this operating range, an encoder rate control should en- sure that no frame exceeds a limit of 50 Mbit/s divided by the number of frames per second. For example, at 29.97 I-frames per second, no frame may have more
27、 4.1.3 Operating ranges 3A and 38 (HDW- Any GoP) Operating range 3A covers HDTV coded at up to 80 MbWs, and may use temporal predictive coding. Operating range 38 covers HDTV coded at up to 175 Mbit/s, and may use temporal predictive coding. (See note below.) B Page 3 of 8 pages RP 213-2001 4.1.4 Op
28、erating range 4 (HDTV - I-only) Operating range 4 covers HDTV coded at up to 300 Mbits, using no temporal predictive coding. For this operating range, an encoder rate control should en- sure that no frame exceeds a limit of 300 Mbis divided by the number of frame$ per second. For example, at 29.97 I
29、-frames per second, no frame shall have more than 10,010,000 bits net data. For HD I-frame-only applications, future equipment designs should conform to this operating range. When coupled with the bitstream constraints defined in 4.4, this will facilitate bitstream interchange over the full range of
30、 MPEG-2 compliant equipment. NOTE - No lower bit rate limit is specified in ISO/IEC 13818-2. Please refertothisstandardforthe precise details. Operating Range 4 HDTV Up to 300 MbiVs I-only coding Operating Range 2 SDTV Up to 50 MbiVs I-only coding 4.1.5 Relationships among operating ranges Relations
31、hips among different operating ranges are illustrated in figure 2. Operating range 2 is a subset of operating ranges 1 and 4. Operating range 1 is a subset of operating ranges 3A and 3B. Operating range 3A is a subset of operating range 3B. 4.2 Compatibility with ISOAEC 13818-2 For all operating ran
32、ges, the MPEG buffer model defined in ISO/IEC 13818-2 (MPEG-2) shall be re- spected. The bit rates specified are all maximum bit rates, and shall include all data that are included in the MPEG-2 buffer (VBV) calculation. Operating Range 3B HDTV Up to 175 Mbit/s Any GOP structure Operating Range 3A H
33、DTV Up to 80 Mbit/s Any GOP structure Operating Range 1 SDTV Up to 50 Mbit/s - Any GOP structure Operating Range 2 SDTV Up to 50 Mbit/s I-only coding Figure 2 - Relationships among operating ranges Page 4 of 8 pages RP 213-2001 For operating ranges 2 and 4 using I-only GoP structure, the video eleme
34、ntary stream is also con- strained such that no frame exceeds the additional limits specified in 4.1.2 and 4.1.4. Devices operating in operating ranges 2 and 4 may have constraints that preclude processing a VBR input directly. For interoperability in the case that a CBR device cannot process a VBR
35、input, that CBR device should pad the incoming VBR signal up to the appropriate maximum bit rate. If any constant bit rate device has insetted such padding, it should remove this padding at its interoperable output interface (see annex Afor further details). NOTE - ISOAEC 13818-2 treats constant bit
36、 rate (CBR) systems to be a constrained version of variable bit rate (VBR) systems; consequently, all systems specified by this practice should be capable of interoperation within a VBR environment. 4.3 Spatial alignment of coded image To ensure optimal muttigeneration picture perorm- ance, the spat
37、ial alignment of coded images specified in SMPTE RP 202 should be applied to both MPEG-2 encoders and decoders. 4.4 MPEG bitstream parameters Operating ranges 2 and 4 are primarily defined to facilitate bitstream interchange within an I-frame-only MPEG-2 VTR (video tape recorder) environment. The ke
38、y requirements of such recorders are listed below: Annex A (informative) Variable bit rate and constant bit rate Some devices have a great deal of flexibility to deal with bursts of data, as might happen when coding a difficult video scene. These devices use variable data rate by increasing data rat
39、es when necessary to preserve quality, and decreas- ing data rates with easier content to improve efficiency. These devices, therefore, are sometimes referred to as providing constant-quality operation. Other devices inherently operate with data rate constrained to a constant value. When data rate i
40、s fixed, there will be some picture quality variation, which will be a function of the picture complexity. If data rates are sufficiently high, these variations can be imperceptible. The ease of processing constant bit rate streams is, therefore, attractive in some applications. - Afixed number of c
41、oded bits (bytes) per I-frame; - To allow random accessfeatures, similar to those available on analog VTRs, some additional con- straints are required as follows: - A defined slice structure; - The repeat of sequence headers on each frame; - The repeat of nondefautt quantization tables on each frame
42、; - In addition, to facilitate the conversion of elementary streams to transport streams, an accurate vbv-delay value should be carried in the video elementary stream. The value of oxmf (Le., variable bit rate) is not allowed. The vbv-delay value may be relied on for remultiplexing elementary stream
43、s into transport streams. SMPTE 356M is an example of an operating range 2 compression system for an MPEG-2 VTR. 5 Audio interchange For audio, no single worldwide compressed standard has been adopted; various transmission systems are in use depending upon geographic area. Global inter- change can,
44、therefore, only be achieved by specify- ing a noncompressed audio format. All equipment should, at minimum, support the use of two AES/EBU audio data streams (four audio channels) of 16-bit minimum resolution plus VUC bits. It is recommended that these audio data streams be limited to 48-kHz samplin
45、g. Regardless of the type of compression (VBR or CBR), all practical systems need some limits on allowable bit rate variations. To address this, MPEG-2 specifies a buffer model for both compression and transport. It is the respon- sibility of the compression encoderto manage the data rate, through v
46、arying quantization granularity, to avoid buffer overflow or underflow. Perhaps the most familiar example of VBR implementation is DVD, where storage efficiency is especially critical. The average bit rate on a DVD is around 4.8 Mbis while the peak rate may be 9.8 Mbis. Professional television equip
47、- ment has used both VBR and CBR; VBR is popular with some disk recorders, while CBR compression has generally been used on tape and in some editing disk recorders. Page 5 of 8 pages RP 213-2001 With clear applications for both VBR and CBR in the profes- sional domain, this practice facilitates an a
48、pproach to inter- connect VBR and CBR components in a system. This approach is based on an interface in which VBR signals are Annex B (informative) Interoperability There is confusion in the minds of many people who use systems about the interoperability between system compo- nents. One of the reaso
49、ns for this is that the degree of interoperability varies in different situations. This is a recommended approach to creating a common language that allows users and providers to understand more clearly what is meant by interoperability. The idea behind the approach is to define four levels of interoperability. These levels bear a resemblance to the OS1 model in that the lower level of interoperability is about interoperation at the physical layer and the link layer, the middle level is about interoperation at the transport layer, and the highest level of interoperation is about int
