The VC-2 video codec is the subject of a number of SMPTE standards and grew out of work at BBC R&D on an open and royalty-free codec called Dirac. Development continues and the standard is currently being updated to add new coding tools and to improve support for UHD video formats.
Project from - present
What we're doing
VC-2 is specified in SMPTE ST 2042-1 and there are related standards and recommended practices in the 2047 series which describe coding parameters for particular applications and carriage of VC-2 compressed video over SDI. We have updated the main standard to support next generation video formats, to add new coding techniques and to remove some old and unused elements. We have also written a recommended practice, RP 2047-5 (amended), to cover compression of UHDTV video to the same bit rate as that of uncompressed HDTV signals so that existing infrastructure can be used for these signals.
Why it matters
TV production is making increasing use of IP networks but the bit-rates needed for uncompressed video, particularly for UHD, can easily exceed the capacity of available links. A relatively mild level of compression (referred to as “mezzanine”) can maintain the picture quality needed for production while enabling much more efficient use of network capacity. The simplicity of implementation and the low latency of the VC-2 codec make it ideal for these applications. The fact that it is royalty-free is a further benefit for manufacturers and users.
How it works
VC-2 is a low-latency video codec that uses wavelet transforms and entropy coding. It can be readily implemented in hardware or software at very high bit rates. It is derived from a subset of the Dirac codec developed by BBC R&D in the early 2000s. A detailed description of the algorithms can be found in BBC R&D White Paper 238.
Several companies are using VC-2 in their products, serving the needs of broadcasters such as the BBC and providing greater choice in the marketplace.