BBC Research & Development has begun a five-year collaboration with academic researchers from the University of Surrey for audio-visual research and the University of Salford for research into acoustics. The benefits of this research will be shared with the industry to enable innovations like the recently launched HD audio and Radio Player, writes David Fox.
BBC R&D will also work closely with the universities of Southampton, Queen Mary University of London (QMUL) and York. All five of the partner universities are world-leaders in audio research.
This forms a new partnership model for BBC R&D and demonstrates a growing commitment to increase its collaborative work with external partners from both academia and industry. It will lead to additional collaborations with other UK and international universities and research institutes, in addition to close working relationships with industry.
“This collaboration is another step towards more innovation in radio. It will deliver significant benefits to the industry and listeners,” said Tim Davie, the BBC’s Director of Audio & Music.
BBC R&D has been doing audio research for decades, ranging from loudspeaker design and acoustics for studios, some of the first work on digital audio, and the development of standards for loudness measurement. “The need for the BBC to be at the forefront of audio and audio-visual R&D has never been greater, with ever-advancing developments in technology, that offer new ways for audiences to consume our audio content,” said Dr Graham Thomas, principal researcher, BBC Research & Development. “There is also a real need to improve efficiency and reduce costs. However, we do not have the in-house resources to do justice to the full range of possibilities that audio now has to offer, and even if we did, we would be missing a trick if we overlooked the world-leading expertise available in the UK academic sector.”
“We are certain that by combining the expertise of world-leading academics with the practical industrial knowledge of BBC R&D engineers we will significantly advance the field of audio research for broadcasting,” added Matthew Postgate, the BBC’s Controller Research & Development.
Professor Yiu Lam, Head of Acoustics Research Centre, University of Salford, hailed it as a “major initiative, which will let us work with a leading media organisation to transform our knowledge in acoustics research into delivering cutting-edge technology.”
Professor Adrian Hilton, head of the University of Surrey’s Visual Media Research Group, has collaborated with the BBC on different projects over the past decade introducing novel technologies for both audio and visual projection, and said: “We hope that our collaboration will continue to flourish and provide significant benefits to those who use watch and listen to the BBC’s broadcasts across a wide spectrum of programmes.”
The new BBC R&D lab at MediaCityUK, Salford, includes a new listening room (pictured with a set of speakers set up for 3D sound tests). The BBC is currently looking to recruit a senior researcher with a strong background in audio, who will be based at MediaCityUK and will lead BBC R&D’s audio team across both its Salford and London labs, as well as leading its involvement in the Audio Research Partnership.
The audio centre will focus on four core areas of research initially:
Source separation (independent component analysis for audio un-mixing – such as to help eliminate microphone cross-talk in recording, or to allow listeners to re-mix a programme to improve the audibility of dialogue, or even to up-mix archived TV programmes from mono or stereo to 5.1 or full spatial audio by analysing the video to place sounds at the right 3D location);
Audio semantics (automated metadata generation to provide new tools for both programme makers and audiences);
Spatial audio (including Ambisonics [as in the picture top where a Soundfield microphone is used for an experiment], Periphony, and studying the options for true 3D sound);
and Room acoustics (including optimising rooms for spatial audio both at the broadcaster and consumer end).
They will also collaborate on pure research into speech recognition (for applications including archives and accessibility – such as improving the intelligibility of compressed audio, particularly for the hearing-impaired) and audio coding (low latency and IP delivery).