THD204 - TMA05 Part One - Digital Transmission
Compression of digital data is usually performed by a hardware/software device (codec) programmed with a mathematical formula, which reduces the size of the data (codes) and, after storage or transmission, reconstructs (decodes) the information. Two of the common standards for video compression are MPEG, introduced by the ISO Motion Picture Experts Group, and H261.
MPEG
There are two common video standards for this a present:
MPEG-1: designed for a transmission rate of 1.5 Mbps specifically for VideoCD (VCD) and CDi media.
MPEG-2: Designed for coding interlaced images at a rate above 4Mbps - uses include Digital TV (Sky Digital use this) and Digital Versatile Disk (DVD).
An MPEG-4 standard is in the final stages of development and release. It is a much more ambitious standard and addresses speech and video synthesis, fractal geometry, computer visualisation, and an artificial intelligence (AI) approach to reconstructing images.
MPEG-1 and 2 are standards, which define the techniques whereby digital video can be compressed by factors varying from 25:1 to 100:1. This is achieved by using the following five different compression techniques.
- 1. Discrete Cosine Transformation: - of a frequency-based transform.
- 2. Quantisation - sometimes known as "lossy compression" is a technique for losing selective information that the human eye will not detect.
- 3. Huffman coding - "lossless" compression utilising code tables based on the statistics of the data to be encoded.
- 4. Motion compensated predictive coding - the differences between the current image and the preceding image are calculated and only the differences are encoded.
- 5. Bi-directional prediction:- some images are predicted by examining the immediately preceding and immediately following images.
H261
Video coding standard published the ITU (International Telecom Union) in 1990. Designed for datarates which are in multiples of 64Kbit/s, sometimes called p x 64Kbits/s (p is in the range of 1 - 30). This codec was designed for use with ISDN.
The coding algorithm is a hybrid of inter-picture prediction, transform coding and motion detection/compensation. Its datarate is able to be set between 40Kbits/s and 2Mbits/s. Inter-picture prediction removes temporal redundancy, the transform coding removes spatial redundancy and motion vectors are used to compensate for motion. Further redundancy is removed from the bitstream by using variable length coding. This standard was originally used for commercial video conferencing and some cable companies for transmissions such as video on demand also support its use.
These standards for digital compression particularly MPEG-2 are an integral part of the technology required to transmit graphics and video into homes and are used by Sky Digital and ITVDigital (formerly OnDigital). The compressed video is suitable for transmission over the Local Loop using a broadband service like ADSL.