BBC develops 'alternative' codec
With wavelets, 'Dirac' lessens lossiness
Just when we think that the codec wars are just about over, the British Broadcasting Corporation (BBC) has released a codec that works differently from other existing systems. In March it issued it to the open source community through Sourceforge under the name of Dirac, named after the eccentric and brilliant British physicist.
This week the BBC said that it has been working on the codec since January last year, and now it is asking for open source help in taking it towards a product. Dirac is different from existing video compression systems, in that it uses wavelet technology.
Wavelets are mathematical functions that cut up data into different frequency components, and then study each component with a resolution matched to its scale. In effect the video is not, at first, compressed but converted into a waveform and sent, the compression is carried out on the waveform that has been created.
Wavelets were developed independently in the fields of mathematics, quantum physics, electrical engineering, and seismic geology, and they probably owe something to Dirac’s work, hence the name. As these fields have begun working together over the last ten years or so, many new wavelet applications have emerged such as image compression, turbulence prediction and earthquake prediction.
The BBC says that already the system gives a two-fold reduction in bit rate over MPEG 2 for high definition video at 1920x1080 pixels and that it will be further optimized for Internet streaming. This type of performance is roughly in line with the Video Codec 9 which Microsoft uses in its Windows Media Player and only slightly less than the H.264 international standard. At the moment the Dirac codec is in the early stages of development, started as a research tool.
An experimental version of the code, written in C++, was released under an Open Source license agreement on 11 March. The BBC said: "A lot remains to be done to convert our promising algorithm and experimental implementation into practical useable code. This includes optimization so that it can decode in real time. Algorithmic enhancements are needed to improve the compression performance still further. The resulting codec needs to be integrated with other parts of a compression system such as players, and interfaced using standard IO formats."
© Copyright 2004 Faultline
Faultline is published by Rethink Research, a London-based publishing and consulting firm. This weekly newsletter is an assessment of the impact of events that have happened each week in the world of digital media. Faultline is where media meets technology. Subscription details here.
Sponsored: Hyper-scale data management