Project: Intelligent Video and Audio Restoration

VideoStar will develop computer based techniques to enable the cost effective restoration of video media with the minimum of human interaction. This in turn will facilitate the greater use of archive video media held in public and private sector TV archives and libraries, national archives and museums, corporate and personal collections. We do not address the transcription or digitization processes (addressed by previous EU projects) but will research and develop tools and methods for the commercial restoration of images and audio from video so that programmes can be enjoyed by a wide audience on current equipment. While we do not plan to develop consumer solutions, it is inevitable that these will follow._x000D_ Modern audiences have rising expectations of media quality, due to the experience of HDTV and the higher resolutions that are provided by Plasma and LCD Panels (up to 1080 lines progressive scan). We will also develop cost-effective methods of ‘re-mastering’ archive video to meet modern requirements for resolution and fidelity to increase the market opportunities and value of older media._x000D_The target markets for restored Video includes the large number of terrestrial digital TV channels now available, satellite and cable TV services, DVD, Podcasting, on-line and Video on-Demand. _x000D_ Videotape recording technology was first demonstrated by the BBC in 1952, since when at least 27 major analogue and 18 digital formats have been developed . The legacy of video on magnetic tape is now enormous. In 2001, the EC ‘PRESTO’ project estimated that European and North American broadcast archives contained about 100 million hours of professional material on videotape – to say nothing of corporate video. Unfortunately, as stated by the US Library of Congress videotape is ‘the single most unsuitable audiovisual format ever used for archive purpose’. In 2005, the EC ‘PrestoSpace’ project identified over 6 million videotapes in 31 archives in Europe as part of the effort to save the video legacy. The findings were alarming: after only 10-20 years, each video format becomes obsolete. The life expectancy of the material itself varies, but without cold, dry storage most audiovisual materials deteriorate after 20 to 30 years. The survey estimated that 40% to 70% of existing material is in danger of disappearing by 2025 . _x000D_Old videotapes suffer from many different problems that degrade the image and sound. Mechanical defects include damaged edges, folded tapes, torn tapes repaired with a mechanical splice, and tape deformation due to non uniform winding. The oxide coating and the tape itself deteriorate, tapes stick together, clog the heads, and create deposits that interfere with the image. Images suffer from a wide range of colour distortions and fades, flicker, judder and a host of artefacts caused by electronic and physical faults. As the quality of the devices used for viewing increases, with the uptake of HDTV receivers, these degradations become even more visible and annoying. Faced with the decay of their video collections, the archives are concentrating on the overwhelming task of saving the analogue video by transferring it to digital formats, so that content can be preserved by repeated duplication without further loss. They do not – cannot – address the problem of restoring the decayed material to a pristine form so that it can be enjoyed by new audiences. _x000D_ Existing restoration tools follow a traditional workflow, which require a skilled operator to perform the analysis: they deal with either audio or video, but not with both; and are extremely labour intensive. These factors severely restrict their commercial usefulness. The combination of R&D expertise from the image and sound sectors will enable VideoStar to develop integrated, automated methods that are both more effective and more affordable._x000D_ VideoStar will develop and promote the commercially viable techniques that are essential to make the audiovisual cultural heritage of the past fifty years accessible to present and future generations. As well as generating business from the sale of restoration tools and technology licenses, VideoStar will create value for the owners of rights to gain value by promoting and licensing their material for present and future viewing, and create business for service providers._x000D_ The project Ps are:_x000D_ Digital Vision (Sweden – Project lead) [DV] a research-performing SME have experience in designing and exploiting hardware and software solutions for film and video image restoration_x000D_ Cube-Tec International (Germany) [CT] a research-performing SME specialising in audio technology and restoration, with experience of previous European projects in conservation and restoration of media._x000D_ Studio Hamburg (Germany) [SH] a major European Facility House providing Film and Video postproduction and restoration services. Studio Hamburg has previous experience in European projects in conservation and restoration of media._x000D_

Acronym VIDEOSTAR (Reference Number: 4304)
Duration 01/06/2009 - 01/06/2011
Project Topic Our project is to develop cost effective and efficient technologies and systems to restore and remaster material currently stored on video media. We plan to develop a commercial system that will restore both audio and video content on one workstation, to meet modern viewer quality expectations.
Project Results
(after finalisation)
In the VideoStar RTD project new computer based techniques was developed to enable the cost effective restoration of video media including the audio tracks with the minimum of human interaction.
Network Eurostars
Call Eurostars Cut-Off 1

Project partner

Number Name Role Country
3 Studio Hamburg Postproduction GmbH Partner Germany
3 Digital Vision AB Coordinator Sweden
3 Cube-Tec International GmbH Partner Germany