Sustainability of Digital Formats: Planning for Library of Congress Collections

According to ISO/IEC 13818-1, MPEG-2 Transport Stream (MPEG-2 TS or MTS) is one of the two (the other being MPEG-2 Program Stream (MPEG-2 PS)) system coding options for MPEG-2 Video Encoding (H.262) video and audio data. Transport streams and program streams are each logically constructed from packetized elementary streams (PES) but differ in their intended use and how they handle multiple programs. Charles Poynton's Digital Video and HD Algorithms And Interfaces (2nd Ed.) (link via Internet Archive) explains more about PES: "An elementary stream is packetized into packets of 188 bytes, the first byte being MPEG’s sync byte valued 47h. Some systems construct 204 byte packets, expecting the channel coder to overwrite the final 16 bytes of each packet; in this case the sync byte will be B8h" (p. 556). ISO/IEC 13818-1 explains that "Both the transport stream and program stream ... provide coding syntax which is necessary and sufficient to synchronize the decoding and presentation of the video and audio information, while ensuring that data buffers in the decoders do not overflow or underflow. Information is coded in the syntax using time stamps concerning the decoding and presentation of coded audio and visual data and time stamps concerning the delivery of the data stream itself. The 2018 edition of the specification published as Recommendation ITU-T H.222 goes on to say that "the transport stream additionally supports the multiplexing of video and audio compressed streams from multiple programs with independent time bases. For almost error-free environments the program stream is generally more appropriate, supporting software processing of program information. The transport stream is more suitable for use in environments where errors are likely." Use cases for transport streams includes environments "such as storage or transmission in lossy or noisy media" including live streaming events, broadcasting TV channels, and other applications where a stable and continuous stream is essential. Poynton (link via Internet Archive) states "a transport stream is suitable for applications where a player connects to a transmission in progress (like television), as opposed to reading a file from its beginning" (p. 556) and that transport streams "are generally appropriate for distribution, but not for acquisition or production" (p. 167).

Transport streams may be either fixed/constant or variable rate, although Wikipedia states they are most often broadcast at CBR (Constant Bit Rate) rather than (VBR) Variable Bit Rate and "and filled with padding bytes when not enough data exists." The transport stream rate is defined by the values and locations of program clock reference (PCR) fields, which in general are separate PCR fields for each program. Understanding Constant Bitrate (CBR) in MPEG-TS Transport Streams expands on this: "The transport stream rate is defined by the values and locations of Program Clock Unlike video encoding, when CBR is referred to in the context of MPEG-TS, it doesn't refer to the video encoding data rate being constant, but rather to the transport stream's overall data rate. The transport stream can carry packets of many different types of data, including video, audio and captions. The bitrate of each data type can vary with time. A CBR MPEG-TS must ensure the total bitrate remains constant throughout the stream. To achieve this, MPEG-TS encoders and multiplexers transmit can "NULL" or "Stuffing" packets, when there is nothing to send, ensuring that the total number of bits per second remains constant. As a result, even if the stream content bitrate fluctuates, the overall transport stream bitrate remains constant by inserting the requisite number of NULL packets. Reference (PCR) fields, which in general are separate PCR fields for each program." Comments welcome

From a technical perspective, a transport stream comprises 188 bytes including a 4-byte or 8-bit field header (whose starting value is '0100 0111' [0x47]), a 13-bit packet identifier (PID) which specifies the type of data stored in the payload including null packets, and maximum 184 bytes of payload.

Widely adopted for filemaking, DVD disks, and other applications. Most significant is the format's required use in digital terrestrial broadcasting to homes in the United States and several other nations, as governed by the ATSC (Advanced Television Systems Committee) specifications. Also widely used in in digital video cameras, recorders, and players and for streaming applications. Many software tools exist for encoding and decoding.

A modified version of MPEG-2 transport stream is used for Blu-ray Discs. See Blu-ray disc Audio-Video (BDAV) MPEG-2 Transport Stream (M2TS).

Technical (coding) information is contained in the MPEG-2 bitstream in macroblock headers, slice headers, picture headers, Group of Picture (GOP) headers, and sequence headers. The sequence header provides data needed before decoding can begin, such as the size of the picture and the frame rate. Because it is so important, the sequence header is usually repeated as often as twice a second.

Accessibility Features

MPEG-2 TS does not have specific features that enhance or support accessibility but the MPEG-2 family overall provides moderate accessibility support. According to the System Implementation of CEA-708 and CEA-608 Closed Captioning and Program-Related Data, "files that use MPEG-2 video compression encoding may carry closed caption data in the video user data, as per ATSC A/53 Part 4". Typically, accessibility features such as captions and subtitles are carried in containers and wrappers, not in encoded video data.

See W3C's Making Audio and Video Media Accessible for more general information about accessible sound and moving image media.

Moderate to very good, given that the options for audio employ lossy compression. Widespread use is made of MPEG-2 layer 3 audio, aka MP3_ENC. Some producers make use of Advanced Audio Coding, aka AAC, considered to be superior to MP3 at a given bit rate.

Producers are not limited to the audio structures inherent in MPEG-2. The Society of Motion Picture and Television Engineers (SMPTE) provides specifications for incorporating AES3_SMPTE data into the MPEG-2 transport stream. Among other things, this permits the use of AC-3 compressed audio (aka Dolby Digital).

For all types of compressed audio, the quality of the outcome will depend on the type and extent of compression, and the encoder used.

Multi-channel sound is supported in both the backwardly compatible (BC) audio structures shared with MPEG-1, which includes layer 3 audio (MP3_ENC), and in the non-backwardly compatible AAC audio structure. MPEG-2 BC can deliver five main channels and an optional LFE (Low Frequency Encoding or Effects) channel, i.e., 5.1 surround sound. AAC provides a capability of up to 48 main audio channels, 16 LFE channels, 16 overdub/multilingual channels, and 16 data streams.

The use of AC-3 compressed audio (aka Dolby Digital), described above in connection with AES3_SMPTE data, provides another means for presenting surround sound.

From Gary Kessler's File Signatures Table. Comment: "This is not a lot to go on, but MPEG-2 Part 1 Transport (MP2T) files are reportedly broken into 188-byte packets and the 0x47 byte is the sync byte, so should repeat every 188 bytes in the file."

See also Pronom fmt/1055 which states the signature follows this pattern: "BOF: offset 4, 0x47 sync byte appearing every 192 bytes from hereon" quoting from US Patent US7106946B1: Transport stream processing device, and associated methodology of generating and aligning source data packets in a physical data structure.