Revert to Section 4.5.4

A DTTB transport architecture should be flexible and capable of supporting a number of audio, video, and data services through its system multiplex. Data services may be program related or non-program related. SMPTE and others have identified program related data services that could be communicated from the program source that would be helpful to the display of the program. The identified functionalities are seen as desirable for use in the receiver to improve the system performance or to enhance the service for the viewer. Some of the functionalities are viewed as useful in distribution networks to support international program exchange, for scalable service environments or for use in a simulcast implementation scenario.

The transport layer syntax allows the definition of a program map which permits identification of individual audio services by their compression algorithm as well as identification of multiple language channels that can be selected by the viewer or by the distribution network. This requirement to identify compression algorithms allows selection of an audio service (monaural, stereo, or surround sound) and bit rate appropriate to the associated program.

A program service can be provided as an ancillary data service with its own PID. This could take the form of a program guide that is personalized by the service provider. The information required can be supported by a low refresh rate that would not consume a significant amount of the channel bandwidth.

Captioning information, like audio associated with the video, must be synchronized with each television frame. Captioning information should be uniquely identified and carried as user data within the video picture layer. However, the value of using PES packets or sections to maintain commonality in processing at the receiver between captioning and other applications should be considered further.

Closed captioning is a captioning service designed for the hearing impaired. Like general captioning information, closed caption services must be synchronized with each television frame and should be uniquely identified and carried as user data within the video picture layer. However, nothing in the MPEG-2 syntax would prevent closed captioning data being sent in a separate PID, and in some applications this might have some advantage over the carrying the data within the video picture layer. The value of using PES packets or sections to maintain commonality in processing at the receiver between captioning and other applications should be considered further.

Program source identification and program identification information has a great many uses. One application is to allow automatic access to programming for recording and delayed playback by the viewer. Program source and program identification should be uniquely identified and carried as an ancillary data service with its own PID.

Conditional access systems can be supported by the transport syntax with bits identified in the packet header. Information about the conditional access information including key information should be uniquely identified and carried as private data.

Some parties interested in implementing DTTB services intend to provide a range of scalable services for use in different reception environments. Compressed and encoded image sequences may also serve as a format for program interchange. The ability of the video syntax to carry the details of the picture sampling structure used in the coded image, including samples per line, lines per frame, frames per second, scanning format (interlace or progressive) and aspect ratio facilitates use of the program material across a broad spectrum of applications.

Information on the colorimetry characteristics of the encoded video can be supported in the video sequence layer. This includes a description of the colour primaries, transfer characteristics, and the colour matrix coefficients, and allows the receiver device to properly accommodate image sequences derived from sources using different colorimetry.

Conventional television receivers will dominate the market at the start of DTTB services and will populate the market for many decades thereafter. The advantages of DTTB services may lead to a desire to make these services available to existing conventional (NTSC, PAL, or SECAM) receivers.

Providing colour field information in the video syntax helps the decoder re-encode the image sequence to a conventional service compatible output with reduced artifacts, particularly when the source image sequences were derived from related program material.

Automatic scene change detection algorithms may be used in some encoders to improve coding efficiency. Such scene change information, when supported by a production facility, could prove useful to the video encoder at both the compression and transport levels. The information could also prove useful to distribution systems to identify points in the data stream where switching between sources of transmitted bit streams could take place.

There is a further requirement to identify points in the transmitted bit stream other than scene changes where switching between sources of transmitted bit streams or where packet replacement can take place without noticeably disrupting the performance of the receiver. These are termed "clean-insertion" points and are useful for down-stream (local, national, or regional) service providers to modify a cooperative or network service to accommodate it for local use.

Systems for use in the 60 Hz environment can be optimized for transmitting film originated image sequences by transmitting the frame rate of the coded bit stream. This allows encoders to maximize coding efficiency by not transmitting redundant fields and signals the decoder the proper order for displaying the decoded pictures. The DTTB frame rate syntax can be supported within the video sequence layer to support frame rates of 23.976 (24 ¸ 1.001), 24, 25, 29.97 (30 ¸ 1.001), 50, 59.94 (60 ¸ 1.001), and 60 Hz as well as an extension for future capabilities.

4:3 aspect ratio receivers will dominate the market at the start of wide-screen (16:9) aspect ratio services and will populate the market for many decades thereafter. The advantages of wide-screen DTTB services may lead to a desire to make these services available to existing analogue based receivers and other 4:3 aspect ratio display devices.

Pan and scan information could be transmitted as an extension of the picture layer syntax. The pan and scan extension would allow decoders to define a rectangular region which may be panned around the entire coded image, and thereby identify a 4:3 aspect ratio window within a 16:9 coded image.

Random entry into the application bit streams such as video and audio is necessary to support functions such as program tuning and program switching. Random entry into an application is possible only if the coding for the elementary bit stream for the application supports this functionality directly. For example, a DTTB video bit stream might support random entry through the concept of Intraframe coding (I-frames that are coded without any prediction, and which can therefore be decoded without any prior information). The beginning of the video sequence header information preceding data for an I-frame could serve as a random entry point into a video elementary bit stream. In general, random entry points should also coincide with the start of PES packets where they are used, e.g. for video and audio. The support for random entry at the transport layer comes from a flag in the adaptation header of the packet that indicates whether the packet contains a random access point for the elementary bit stream. In addition, the data payload of packets that are random access points also start with the data that forms the random access points into the elementary bit stream itself. This approach allows the discarding of packets directly at the transport layer when switching channels and searching for a resynchronization point in the transport bit stream and also simplifies the search for the random access point in the elementary bit stream once transport level resynchronization is achieved.

A general objective is to have random entry points into the programs as frequently as possible, to enable rapid channel switching.

This functionality is important for down stream switching of packets (inserting local programming such as public service messages or commercials) into an existing bit stream. In general, there are only certain fixed points in the elementary bit streams at which program insertion is allowed. The local insertion point has to be a random entry point but not all random entry points are suitable for program insertion. For example, in addition to being a random entry point, the VBV_delay (video buffer verifier delay) needs to be at a certain system defined level to permit local program insertion. The VBV_delay information can be computed and transmitted as part of the header data for a picture in the compressed video stream. It thereby defines how full the decoder video buffer needs to be before the bits of the current picture are extracted from the buffer and synchronizes the encoder and decoder processes. This is required to control the memory needed at the decoder for buffering data and to prevent buffer overflow or underflow. Local program insertion also always takes place at the transport packet layer, where the data stream splice points are packet aligned. Implementation of the program insertion process by the broadcaster is aided by the use of a splice_countdown field in the adaptation header that indicates ahead of time the number of packets to countdown until the packet after which splicing and local program insertion is possible. The insertion of local programming usually results in a discontinuity in the values of the PCR received at the decoder. Since this change in PCR is completely unexpected (change in PCR values are usually only expected during program change) the decoder clock could be thrown completely out of synchronization. To prevent this from happening, information is transmitted in the adaptation header of the first packet after the splicing point to notify the decoder of the change of PCR values (so that it can change the clock phase directly instead of attempting to modify the clock rate). In addition, there are constraints on 1) the length of the bit stream that is to be spliced in, to assure that the buffer occupancies at the decoder both with and without the splice would be consistent, and 2) the initial VBV value assumed when encoding the bit stream to be spliced in, in order to prevent decoder buffer underflow or overflow.

In broadcasting services, two essential functions are necessary: they are, the function of receiving a certain broadcast channel continuously without any action, and the function of automatic reception or recording of an individual programme, simultaneously. Therefore, it is necessary to define a new descriptor, which is called "Event Descriptor", to identify the individual programme, because the programme number corresponds to the programme channel. An example for the descriptor is shown in Fig. 44.

figure 44

In the MPEG-2 systems, each programme can be received only after the PAT and PMT are received, and some delay occurs when one selects or changes the channel. In order to minimize this delay, a self-cross indicator is introduced into the PAT or NIT.

It indicates whether the PAT or NIT is for the information of the transport stream from which a programme is viewed, or for the other transport streams (channels) which can be received. By this function, the information of other channels (streams) can be obtained while viewing some programme, and it provides assistance for channel selection.

Continue to Section 5

Return to DTTB Tutorial Table Of Contents

Return to Tutorial Index Page