The restoration is made available on our YouTube page** and for theatrical projection. The restored digital version is saved in addition to the preserved film, and is classified as a restoration that is a derivative of the preserved record. The preserved record consists of reels of film protected in cold storage. In the case of director James Blue’s The March, the Lab photochemically preserved the film before it was ever scanned for the restoration. After subsequent “restorations”, the digital version becomes more and more difficult to connect to the image as it actually existed. Digital restoration tools designed to work on film images may be confused by the digital manipulations of the first restoration.
![digital film restoration digital film restoration](https://img.yumpu.com/44672201/1/500x640/digital-film-restoration-in-krems-europa-cinemas.jpg)
However, if we save only the restored version of a film today, the results we are able to achieve in ten years from that starting point will only add further distortions to the visual and audio information contained in the film.
DIGITAL FILM RESTORATION SOFTWARE
Because motion picture restoration software is rapidly improving, the restoration we are able to perform in 2024 will be better than the restoration we do in 2014. The preserved copy of the film can then act as the best possible starting point for future restoration efforts. When we restore a film in the Lab, we always make sure that it has been preserved first. There is no way to return the image to its original state. ” Restoring a digital image in a way that permanently alters its pixels breaks this rule because these changes cannot be undone. In the National Archives Motion Picture Preservation Lab, we try to live by the “Rule of Reversibility”*, which states, “To the degree possible, a conservator should not undertake any procedure or treatment that he or she cannot later, if necessary, undo without harm to the. You may wonder why it matters that the image has changed if it looks better in the restored version. In the Motion Picture Preservation Lab, we use specially designed software to digitally restore films like The March (1963) and John Huston’s Let There Be Light (1946). The image looks better, but it is no longer the same as in the original film. In the digital realm, using specially designed software, we are able to alter the image at the pixel level so that many imperfections disappear. When you watch a film that is scratched or covered with blotches from water damage, these imperfections may distract you from what is happening onscreen. If we say we have restored a film, we are saying that we have enhanced the film for exhibition. Successful preservation means that the image we have now is the same as the image we had yesterday, last month, or even decades ago. Preservation actions include storing a film properly in an inert can in cold storage, but they can also include copying the image from a deteriorated film base to a stable one. If we say that we have preserved a film, we are saying that we have taken steps to protect the integrity and accessibility of the images contained on the film.
![digital film restoration digital film restoration](https://gammaraydigital.com/sites/default/files/2020-07/Guston01_UNRESTORED_NEW.jpg)
The differences between preservation and restoration are subtle, but in the archives world they are profound. The problematic of generating good training data for the particular problem to solve will be shown and possible solutions will be presented.When you watch NARA’s video for The March on YouTube the first thing you see onscreen is a note that the film was “Preserved and Restored by the National Archives.” You may wonder why we make the distinction between preservation and restoration. We will discuss the advantage and disadvantage of the AI approach versus the classical approach in digital film restoration. In the presentation we will show current results of the DeepRestore prototype and we will compare the speed and quality of the restoration with the classic dust & scratch filters from the DIAMANT-Film Restoration Software. A focus is on applying modern machine learning techniques, in particular convolutional neural networks, to remove dust and scratches in archival footage. The goal of the project is to evaluate AI technologies in the sector of digital film restoration. (Austria), running from 01/2018 until end of 2020.
![digital film restoration digital film restoration](https://i.vimeocdn.com/video/834206257_1280x720.jpg)
Presented at JTS2019 by Franz Hoeller at the Netherlands Institute for Sound and Vision, Hilversum (NL) on Saturday, October 5, 2019.ĪBSTRACT: DeepRestore is a research project driven by HS-ART Digital (Austria) together with the TU-Graz Institute of Computer Graphics and Vision.