Navigation auf uzh.ch

Suche

Seminar für Filmwissenschaft

AFRESA: Automatisches Filmrestaurierungs- und Erfassungssystem für Archive.

Prof. Dr. Barbara Flückiger (Leitung), Dr. Franziska Heller
Laufzeit: Oktober 2008 bis April 2011

AFRESA is a joint venture research project, funded by the CTI (the Swiss Confederation’s Innovation Promotion Agency), the University of Basel, the University of Zurich, the Zurich University of the Arts and the film technology industry – namely Swiss Effects Film GmbH and Sondor Willy Hungerbühler AG.

Objective

AFRESA is developing an integrated system for the digitization, preliminary restoration and automatic registration of archival films. The system will enable small (Swiss) archives to perform low-cost, in-house digitization at high quality. Thus the archives can conduct and monitor the digitization process by themselves. The system consists of a scanner, reconstruction software, and a registration system based on the automatic extraction of metadata. The team has also created a manual for preservation issues and guidelines. Lastly, a tool for visual inspection completes the system.

The tagline of AFRESA is: It has to be simple. The initial technological idea behind AFRESA is the integration of all components necessary for digital film restoration in order to register and catalog existing film material. Currently all the different components – e.g. scanner or restoration software – are available on the market but only independently from each other and cannot be used in a particularly automated way. AFRESA aligns all the elements. It aims at being easy to handle for people working in archives, as well as at being as efficient and as affordable as possible. 

Project Partners

The interdisciplinary research of the co-operation partners is divided into the fields of device development, software development and historical research as well as cultural analysis.

The industry partners are responsible for the set-up, adjustment and calibration of the scanner. Sondor, using its experience in the field, has set up a transport mechanism that is suitable for fragile archival film stock. The company has developed an application based on an existing mechanism originating in the audio domain. They will also provide maintenance and service.

Swiss Effects is testing the software with regard to practical matters. They are mapping the scanning alignment with different film stocks. For reasons of time efficiency, the filmstrip constantly moves through the scanner. Very short, non-damaging flashes from LEDs capture single frames. The resolution is > 2K.

The Imaging and Media Lab has specialized for years in the reconstruction of faded color and the repair of other defects in damaged photographs as well as in the digitization of heterogeneous photochemical material. Additionally they are working on solutions for long-term storage. On the basis of the extraordinary achievements of the last decades, the Imaging and Media Lab is now developing the software that facilitates the easy, automated reconstruction of damaged film images. Its algorithms are based on the analysis of the physical and chemical properties of the different film stocks. IML will also develop a database where the reconstructed films are registered and the associated data is recorded and documented. For this purpose IML is working on a method for the automated extraction of inter-titles and texts.

The Institute for Cinema Studies of the University of Zurich is working on the sensitive cultural, ethical and aesthetic issues of the digitization process. The Institute has to ensure that preservation guidelines are respected throughout the entire workflow. For this purpose the researchers are networking with experts in many European countries where huge digitization projects are planned or already in operation to guarantee the interoperability of the system. The film scholars are developing a manual that aims at setting up guidelines for the process. It is important to keep in mind that digitization, which frames the way the films will be seen in the future, is only a snapshot in time related to specific institutional, economic and aesthetic interests. It is deeply linked to the technological state of the art that can change rapidly. Therefore it is crucial to document these surrounding discursive elements by considering the social and cultural contexts of the digitization process.