mpeg - MPEG compression format
MPEG is a set of standards for digitally coding audio and video. The
name is derived from Moving Pictures Experts Group, which is the
technical committee of ISO/IEC (the International Organisation for
Standardisation and the International Electrotechnical Commission)
responsible for developing these specifications.
Several standards have been developed, targeted for different
applications. MPEG-1 encodes non-interlaced material and is optimized
for single-speed CD-ROM bitrates (about 1.5 Mbps). MPEG-2 handles
interlacing and is intended for applications at higher bitrates (4 Mbps
or greater) like studio-quality TV. MPEG-4 is in its early stages and is
targeted for very low bitrates that may be useful for videophone and
other applications. (MPEG-3 was intended for HDTV/ATV but was merged with
Below we describe the components of MPEG-1 (unless otherwise indicated,
subsequent references to MPEG in this document apply to MPEG-1 only).
This standard specifies the syntax for three types of bitstreams :
video, audio, and systems.
Video Bitstream [Toc] [Back]
Compression utilizes tranform coding and motion estimation, which attempt
to remove spatial and temporal redundancies in the original image
sequence. Image dimensions up to 4096 by 4096 pixels and various frame
rates up to 60 Hz are specified (field information is not defined in the
video bitstream), with bitrates up to 105 Mbps allowed.
Audio Bitstream [Toc] [Back]
Compression is based on subband coding, which divides the input into
different frequency bands and allocates bits across frequencies based on
perceptual importance. Mono and stereo sources are supported at sampling
rates of 32, 44.1 and 48 KHz, and allowable bitrates range from 32 to 448
Three layers of encoding are specified : Layers I, II, and III.
Generally speaking, the higher layers provide better compression (lower
bitrates for the same audio quality) at the expense of greater
computational complexity. The higher layers are also supersets of the
lower layers in the sense that a Layer III decoder must be able to decode
an audio bitstream of any layer, and a Layer II decoder must be able to
decode bitstreams from Layers I and II.
Systems Bitstream [Toc] [Back]
The systems specification defines how the compressed audio and compressed
video bitstream are multiplexed; no additional compression is performed.
This layer serves to packetize and interleave the compressed audio and
video data, along with timestamp information and decoder buffering
requirements. Up to 16 video and 32 audio streams may be multiplexed in
a single systems stream.
The MPEG compression format is supported by several developer and enduser
products. Current support is limited to MPEG-1, Layers I/II audio,
and systems streams with one audio track and one video track.
Compression Library [Toc] [Back]
Provides MPEG-1 video and audio encoding and decoding. See
CLintro(3dm), cl_mpeg1(3dm), and cl_aware(3dm).
Audio File Library
Provides MPEG-1 audio encoding and decoding. See AFintro(3dm)
Movie Library [Toc] [Back]
Provides reading and playback of MPEG-1 systems or video
bitstreams. See mvIntro(3dm).
Digital Media Tools [Toc] [Back]
Several tools enable end-users to create, manipulate, and play
MPEG-1 bitstreams. Please refer to the man pages listed in the
- sample MPEG-1 systems bitstream file (movie including
The offical MPEG-1 specification :
ISO/IEC International Standard 11172. Coding of Moving Pictures and
Associated Audio for Digital Storage Media Up to About 1.5 Mbit/s.
mediaconvert(1), dmconvert(1), mediaplayer(1), mediamaker(1), CLintro(3),
cl_mpeg1(3), cl_aware(3dm), AFintro(3dm), AFaware(3dm), mvIntro(3mv),
PPPPaaaaggggeeee 2222 [ Back ]