Democratising REsearch and restoration of Audiovisual Media and Film (DREAM-Film): practices and methods to improve the accessibility to restoration and promote competency exchange.

Project Information

DREAM-Film

Grant agreement ID: 101105533

DOI

10.3030/101105533

Project closed

EC signature date 26 April 2023

Start date 1 September 2023

End date 28 February 2025

Funded under

• Marie Skłodowska-Curie Actions (MSCA)

Total cost

No data

EU contribution

€ 140 718,24

Coordinated by

UNIVERSITEIT VAN AMSTERDAM
Netherlands

Periodic Reporting for period 1 - DREAM-Film (Democratising REsearch and restoration of Audiovisual Media and Film (DREAM-Film): practices and methods to improve the accessibility to restoration and promote competency exchange.)

Reporting period: 2023-09-01 to 2025-02-28

Summary of the context and overall objectives of the project

Unlike other cultural heritage materials, film preservation and restoration has received limited technical and scientific research, leaving many aspects of film materials and color processes largely unexplored. While digital image processing offers advanced methods for detail and color reconstruction, film restoration remains highly dependent on manual intervention. Most available tools, such as semi-automatic cleaning software, are proprietary, and color reconstruction is typically performed using video editing software, often without objective reference to the original materials. The restoration process requires specialized personnel, continuous supervision, and costly equipment. This situation makes traditional film digitization and restoration highly resource-intensive, requiring significant human effort and financial investment. Furthermore, the quality of the results often depends on the expertise of the institutions and the funds available, creating a disparity in access to these processes. This issue exacerbates the gap between institutional and non-institutional archives, as well as between archives in different geographical regions. While wealthier institutions can extensively digitize, restore, and promote their audiovisual heritage, smaller archives and those in less developed areas struggle to do so, risking the invisibility of their cultural heritage. This disparity not only threatens the survival of less-documented audiovisual materials but also distorts the historical, cultural, and societal representation of different communities. The resulting imbalance hinders a comprehensive and diverse understanding of archival collections, limiting access to a pluralistic and inclusive view of cultural heritage.
The DREAM-Film project (Democratising REsearch and Restoration of Audiovisual Media and Film) seeks to develop alternative approaches to film materials analysis, digitisation, restoration, and quality assessment. By drawing on the expertise of both well-established institutional archives and smaller non-institutional centres, the project aims to make these processes more democratic and inclusive, reducing restoration costs in terms of time and resources.
Collaborating with the University of Amsterdam (UvA), the University of Milan (UniMi), the Eye Filmmuseum of Amsterdam (EYE), and the Film Center of Sarajevo (FCS), the project fosters an interdisciplinary approach to developing cost-effective and high-quality film restoration solutions. Through this research, the project contributes to open-access knowledge sharing, ensuring that restoration practices are globally inclusive and sustainable.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The DREAM-Film project initially focused on film digitization, color reconstruction, and quality assessment. However, during the research, a gap in scientific studies on film materials emerged, leading to the addition of materials assessment as a key topic. This expanded approach integrated material science with innovative restoration methodologies, enabling a comprehensive evaluation of film scanners and a comparative analysis of digital acquisitions with transmittance measurements from both pointwise and imaging-based techniques.
One of the most significant breakthroughs was the first application of hyperspectral imaging on cinematographic images, contributing to the development of new methodologies for processing multidimensional data and identify film dyes. These studies encompassed both modern and early cinema film stocks, highlighting critical differences in digital acquisition between historical and contemporary materials. The use of pointwise spectrophotometry on early cinema films not only set new research directions for identifying dyes used in tinting, toning, and hand coloring but also paved the way for developing new digital restoration procedures based on the physical properties of the films. By integrating material analysis into digitization and restoration workflows, this research has enhanced archival practices and opened new avenues for film preservation, potentially leading to the creation of the first databases of film dyes in the future.
In the field of digitization, several studies have been conducted to evaluate the performance of cinematic scanners, focusing on the processes used by software for color encoding in the transition from analog to digital, as well as assessing their dynamic range. This research led to the development of preliminary guidelines for scanning cinematographic materials, providing insights into best practices for optimizing digital acquisition. In the future, these efforts could evolve into the definition of standardized criteria for assessing the quality of film scanning systems, contributing to the establishment of industry-wide benchmarks for archival and restoration practices.
In the field of color reconstruction, studies on digital color reproduction from film scanners have led to the development of initial color restoration methods based on the original materials. This research has primarily focused on reconstructing the colors of tinted films but has also explored image inversion processes. Future developments aim at creating an algorithm to simulate the optical printing process, though further data collection on cinematographic materials is still required.
For quality assessment, specific metrics have been defined to evaluate the dynamic range and the presence of glare in optical systems. Looking ahead, these metrics are expected to evolve into a comprehensive framework for assessing the quality of film scanners, providing a standardized approach for evaluating digitization systems.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The DREAM-Film project has introduced significant advancements in film preservation. A key breakthrough was the first application of hyperspectral imaging on cinematographic images, enabling multidimensional data processing for the identification of film dyes. This, along with the use of pointwise spectrophotometry, has not only improved the understanding of historical colorization techniques but also raised the awareness on the development of a comprehensive database of film dyes, a valuable tool for archives and restorers.
A major outcome of the project is the establishment of preliminary solutions and best practices for scanning, ensuring consistency across different archives. The extensive evaluations of cinematic scanner performance, analyzing color encoding processes and dynamic range, led to the development of preliminary scanning guidelines. These efforts could evolve into formal quality standards for film digitization, supporting best practices in the industry, and minimize variations caused by different equipment, improving the reliability and reproducibility of digital acquisitions.
In the field of color reconstruction, the project has informed new restoration methodologies based on the original physical properties of films. Future developments aim at simulating the optical printing process, though further material data collection is needed. The project has also developed advanced quality assessment metrics to evaluate acquisition systems, digitization accuracy, and color fidelity. These tools provide the research community with quantifiable criteria to compare restoration results, facilitating more informed decision-making and enhancing transparency in restoration processes.
To further democratize knowledge in audiovisual heritage preservation, DREAM-Film has produced open-access publications and materials, offering educational resources to support archivists, researchers, and restoration professionals. To maximize the impact of these advancements, further research is needed, as well as standardization efforts and perseverant collaborations between archives, restoration experts, and scanner manufacturers.

infographic