Barb outlines interactive investigations 12 November 2004 Barb has this week outlined some of the steps it has been taking on interactive measurement. The Broadcasters’ Audience Research Board’s Research Director, Tony Wearn, told a conference in Berlin that techniques are being investigated to determine which has the best potential for the future. Challenges are being faced in the UK ahead of many other territories and Barb is at the forefront of development for reporting of interactive services. Speaking at the ASI 2004 European Television Symposium session entitled The Challenge of Interactivity, Tony Wearn said that Barb currently provides data on usage of interactive applications from the digital satellite platform, where these services are separately identifiable as part of the broadcast (using a specific Service Information code). Barb’s current data shows that interactive applications are commonly used and that there is a dramatic increase in the use of interactivity when tied into a particular event, such as Wimbledon, Euro 2004 or The Olympics. To extend current reporting, Barb has been looking at other detection methods that could be introduced, said Tony Wearn. The main techniques being investigated include identifying infrared activity to tag a minute as ‘interactive time’, the extension of the use of audio detection (currently part of the range of techniques available to Barb for identification of TV content) for some interactive applications, and two possibilities of inserting codes into the visual output – visual watermarking and visual barcoding. Most detail was given about barcoding, for which a prototype technique has been developed in the UK by BBC Research & Development. This would allow for identification of a unique barcode which could be inserted on each interactive application. Tony Wearn said: “Fresh thinking needs to be applied to find appropriate technical solutions to extend the capture of interactive viewing. The techniques we are investigating need to be tested for the amount of information that could be captured, the main type of data that could be expected as output, including at what level it would be statistically appropriate to report and ultimately whether it will deliver to the industry something of value.” “The next steps for Barb are to finalise our investigations, consolidate our conclusions on the feasibility of each technique and test against what is viable for the market.”