Photo of Mark Mann

I joined BBC R&D in September 2010 having always found the technology involved in broadcasting fascinating. I was enjoying my post-doctorate job in Electrical Engineering in Cambridge, but developing new technology which has the potential to be used by every household in the country was one of the many things which attracted me to this job.

Though my background was research in Electrical Engineering (PhD) and Physics (MSci), my specialism was Nanotechnology, which meant I lacked some of the core computer programming skills required for the role. Also, before starting the job, I had little exposure to broadcasting. The training provided is excellent. The Broadcast Engineering Foundation courses at Wood Norton were a really useful introduction to the field and I have been sent on programming courses which have helped me greatly with my first two projects.

My first project looked into whether it was possible to determine the mood of a television programme from the music contained within it. Why? The BBC wants to put its entire archive online with a million hours of content dating back to before World War II. There are some great programmes in there which people today won’t know about, so navigating the archive by the mood of the programme will enable people to stumble across something that they might really enjoy. Obviously, it is cheaper to get a computer to ‘tag’ the mood of the programme than a real person, so I’ve used a mixture of signal processing to work out what music is being played and machine learning so that the computer can automatically determine the mood of the programme. During the project I collaborated with Salford University, helped to setup an online music survey, 'Musical Moods', to get some ground truth data and helped to exhibit the project at the Big Bang Fair.

I’m currently working on graphics and sports analysis, with a particular focus on Olympic diving. Here, the C++ programming course I took is really bearing fruit, because this is Photoshop from first principles. The task involves separating the diver from the background during his or her dive to provide a clearer representation of how the diver moves and to extract some biometric data for the commentary team. As part of this project, I am working with BBC Sport and RedBee, who put all the BBC’s programmes on air.

As a trainee you are encouraged to broaden your experience of all aspects of production so that you may understand how a programme is put together to give you a better idea of how improvements and new technology can be implemented. Consequently, I have gone to the recording of programmes such as A Question of Sport, Shooting Stars and Mock the Week. You can also take part in wider BBC initiatives such as School Report, where you go into a school to help pupils to put a news bulletin together. I took a lesson on broadcast technology and taught them how to use software to edit their reports.

All activity (7)