Scenario2-Paintings

Scenario 2
Change of Colours in Paintings and Works on Paper

The Scenario 2 faces the problem of colour change in paintings and works on paper. The main case studies considered in this Scenario are: two versions of the Scream, “The Scream” 1910? (MM.M0514, oil and tempera on cardboard) and “The Scream” 1895 (G0193-03, hand-coloured print on paper) by Edvard Munch, MUNCH Museum in Oslo; “Road in Provence” by Paul Cézanne, a watercolour on paper (inv. no. 1933. 1221, Art Institute of Chicago).

The two version of “The Scream” display color change (fading, darkening) and need special attention for long term preservation. The fading of areas with cadmium yellow visible in “The Scream” (1910?) has been studied in depth and proven to be a result from the interaction of the paint materials with high moisture and/or light in combination with chlorine compounds. Other discoloration phenomena were also observed after the painting has been recovered from a robbery in 2006, a big halo being visible in the left lower corner of it suggesting the action of a liquid agent that damaged both the cardboard and the paint layers.

Watercolours are executed using pigments bound with plants gums. The lean nature of the binder makes pigment particles exposed to a variety of agents of deterioration, including light, pollutants and changes in humidity. As a result of exposure, colour change may occur. “Road in Provence” by Paul Cézanne is being studied to evaluate the colour changes that may have occurred through time.

PERCEIVE will also include the testing of the tools developed and applied to the MUNCH and AIC case studies, on mural paintings affected by colour change and the presence of lacuna, focussing on two specific case studies. As part of PERCEIVE Scenario 1 on lost polychromy, the complex of the Iseum in Pompeii, with its sculptures has been identified as a potential case study since it includes the problem of colour reconstruction and visualization of analytical data for its statues. This site, with its mural paintings, is particularly relevant also for Scenario 2 as it addresses key research challenges as:

The problem of color reconstruction
The rendering of diagnostic data
The color changes over time
The complex task of frescoes recontextualization
Recompositing and reconstruct digitally the missing elements (lacunae) of patterns.

Using archaeological data PERCEIVE will explore the digital restoration and recontextualization of the South portico frescoes. The project will integrate traditional computer graphics with AI and Image-Based Rendering (IBR) to create digitally restored hypothetical versions.

Lastly, a series of frescoes from the Villa of the Papyri in Herculaneum has been also included as case study due to widespread color transformation phenomena caused by the eruption of Vesuvius in 79 AD. The high temperatures from the pyroclastic flow altered the chemical and physical structure of the original wall pigments, resulting in a visible and pervasive shift from yellow to red across the surfaces.

This color change is particularly evident in some frescoes now exhibited in the MANN collection. The objective of PERCEIVE is to use archaeometry to achieve a precise and accurate result useful to digital color reconstruction. By systematically analyzing the altered pigments it’d be possible to determine the original color palette and consequently restore them digitally, till their authentic appearance, thereby preserving their historical and artistic integrity.

From our research questions to our main results

The main questions that researchers in Perceive aim to answer are:

Which colours are mainly subject to change over time and to which extent?
How have they been applied by the artist and how the technique affected their change over time?
What does it mean for a contemporary observer perceive these colours and their change?
How can we predict further change and simulate the rate of damage over time?
Are computational tools available or feasible for such modelling?
How to better transfer to wider audience the knowledge and tools available for the study, monitoring and modelling of colour change phenomenon?
What kind of experiences (digital, immersive, hybrid) can be designed to enhance perception, authenticity and sense of care of colored collections?

The PERCEIVE project takes further the studies around colour changes and assessment and/or prediction of change for the future, by applying digital and machine learning-based methods to large experimental datasets from discoloured laboratory samples (mock-ups), and sensitive art collections as the one from MUNCH.

Based on the practical need of curators and conservators to use “Predictive Systems” for simulating colour changes back and forth for important, sensitive items in the museum collections, the reconstruction of how these changing colours would have been originally perceived is crucial, since their appearance strongly influences our perception of the artworks nowadays and also in the future.

These tools or services could be of high interest also for the citizens, who, while playing the role of scientists, could at the same time perceive the digital reconstruction as more authentic, and, ultimately, developing a sense of care for these collections. It gives the possibility to “play forward the clock” to see how the artwork would look in a span of time if no specific preventive conservation measures are put in place.

Based on acquired analytical and imaging data a specific methodological protocol has been designed using Image Processing Methods (IBM) and Machine Learning approaches. All input data acquired in different analytical campaigns with different instruments, photographs from museum archives, video materials and published papers or conference communications are also available in the Colour Knowledge Repository (paintings and works on paper community)

The input data in form of images and colorimetric measurements (MFT campaigns) led to the development of a LDE (Light Damage Estimator), a smart, user-friendly tool that helps estimate, predict and simulate how an object might change over time when exposed to light, enabling you to identify potential risks before any real damage is likely to occur

The visualization tool Mulax has also been designed to display different type of analytical results from imaging, XRF spectrometry and hyperspectral investigation providing accessibility and interoperability with modern web standards. Through its connection with the PERCEIVE cloud, it facilitates remote data management and processing, enhancing collaboration among researchers.

CKR – Colour change in painted objects community and a selection of contents

LDE Tool – The Scream (detail)

MULAX AVS – Discovery layers for “The Scream” case study

The technology is based on ATON Framework and enables to see exactly what types of analytical data exist for each object and visualise them directly on the artwork. This creates an intuitive way to present the various data modalities and to allow the user or expert to explore them. These visualizations and replicas can help both researchers and the public understand the raw scientific data which needed is considered for each artwork. This underpins the attempted contribution of PERCEIVE, which is to digitally restore the original appearance of the art-object, the evolution of its appearance over time and estimate possible evolution of the change in the future, thus creating a sort of virtual time travel experience.

Two different care prototypes have been designed and showcased as immersive experiences taking input from the scientific studies around the Scream (1910?)

Scream Time Machine – a demonstrator of a multisensory exploration of the history and colour change of this iconic painting, involving a 6.30 min video tutorial, 3D prints of details with colour fading, a 2D full size scan and a board game (June 2024);
Tiny Conservators – a game-based experience for families with kids aged 6 to 10y. This prototype is an interactive game, designed for children and families to explore how colors in paintings change over time, transforming conservation science into a gaming adventure – blending research, storytelling, and gameplay (December 2025).