Art · 6 min read
How Conservators Identified and Removed Overpaint from the Ghent Altarpiece
Technical imaging and staged cleaning revealed preserved Van Eyck paint while leaving ethical judgments visible.

How Conservators Identified and Removed Overpaint from the Ghent Altarpiece
When conservators began a new campaign on the Ghent Altarpiece in 2012, they were not simply cleaning a dirty painting. The large polyptych attributed to Hubert and Jan van Eyck had accumulated varnish, repairs, and substantial passages of later paint. Some additions had remained in place for centuries and had become familiar to viewers. Removing them required evidence that they were not original, evidence that the original layer survived beneath them, and a treatment method that would not sacrifice that layer.
The project shows conservation as a sequence of testable judgments. Imaging can distinguish materials and revisions, microscopic examination can locate boundaries, and tiny cleaning tests can establish whether layers separate safely. Yet instruments do not decide what a work should look like. Conservators, art historians, scientists, church custodians, and public authorities still have to interpret the evidence and accept responsibility for intervention.
A work with a long material history
The altarpiece was presented in 1432 for a chapel in what is now St Bavo's Cathedral in Ghent. An inscription attributes its beginning to Hubert van Eyck and its completion to his brother Jan, although the precise division of labor remains a scholarly question. Its painted oak panels open and close to show different ensembles of figures and scenes, with the Adoration of the Mystic Lamb at the center of the open lower register.
Over nearly six centuries, the panels were moved, stolen, separated, exposed to changing environments, cleaned, varnished, and repainted. A 1934 theft left the panel of the Just Judges missing; the panel on display is a copy. Earlier restorers worked with the materials, knowledge, and aesthetic priorities of their own periods. Their additions became part of the object's history, even when they obscured fifteenth-century paint.
The Royal Institute for Cultural Heritage in Belgium, known as KIK-IRPA, coordinated the current conservation and research campaign with the cathedral, museums, universities, and funders. The treatment proceeded in phases: the exterior panels, the lower register of the open altarpiece, and then the upper register. Work was conducted in view of museum visitors for portions of the campaign and documented through the Closer to Van Eyck website.
Imaging made a layer map before removal
Conservators began with high-resolution visible-light photography, infrared imaging, infrared reflectography, X-radiography, ultraviolet examination, and magnified inspection. These methods respond differently to pigments, binders, underdrawing, damages, and varnishes. Used together, they can show whether a form belongs to the design beneath the surface or to a later intervention.
X-radiographs register materials that absorb X-rays strongly and can reveal construction, nails, dense pigments, and changes beneath the visible surface. Infrared reflectography can penetrate some paint layers and make carbon-rich underdrawing visible. Ultraviolet illumination can make varnishes and retouches fluoresce differently. None supplies a self-explanatory picture. Specialists compare the images with cross-sections, documentary history, known material behavior, and neighboring passages.
This multi-modal evidence supported a striking estimate: large areas of the visible surface had been covered by overpaint, much of it probably dating from the sixteenth century. The overpaint was not confined to small repairs. It altered draperies, architecture, vegetation, sky, and the central Lamb. Crucially, tests indicated that much original Van Eyck paint remained in relatively good condition below.
Removal proceeded under magnification
A later layer can sometimes be removed because its solubility or physical structure differs from the layer beneath it. Conservators develop solvent mixtures or mechanical techniques through small tests, then work slowly under a microscope. The aim is not to strip the panel to bare wood or to make it look newly painted. It is to stop at a chosen historical layer while preserving fragile original material.
The boundary is not equally clear everywhere. Old retouching may fill a loss in one place and extend over intact paint in another. Varnish can be unevenly oxidized. Pigments can have changed chemically. Conservators document each stage so a decision can be reviewed and, where possible, later treatment can distinguish new retouching from original paint.
After removal and stabilization, areas of actual loss may be filled and visually integrated. Contemporary conservation practice generally favors retouching that helps the image read at normal viewing distance but remains identifiable under close or technical examination. Materials are selected with future reversibility and aging in mind, though no treatment is literally consequence-free.
The Lamb made the judgment visible
The restored face of the central Lamb drew public attention because it differed sharply from the familiar one. The later face looked more naturalistically ovine and turned partly to the side. Beneath it, conservators found a smaller, frontal face with large forward-looking eyes and a more explicitly human expression. KIK-IRPA's documentation reports that the original was well preserved enough to recover and that the overpaint extended beyond areas of loss.
The result was initially described in some coverage as strange or surprising. That reaction illustrates a conservation problem: visual familiarity can be mistaken for authenticity. Viewers had known the later layer through reproductions, scholarship, and worship. Removing it did not uncover an image outside history; it selected the fifteenth-century composition over a long-standing sixteenth-century modification.
Technical evidence made that selection defensible. Layer relationships, material differences, and continuity with surrounding original paint supported attribution of the frontal Lamb to the Van Eyck campaign. The decision also fit the project's stated goal of revealing preserved original paint where safe. Still, a different conservation philosophy might retain later additions as evidence of reception. Scientific identification narrows the options; it does not eliminate the value judgment.
Public documentation changed the case
Closer to Van Eyck provides zoomable photographs before, during, and after treatment, alongside infrared and X-ray images. That access allows outside scholars to test claims at a level previously available mainly to the treatment team. It also exposes uncertainty: inscriptions, underdrawing, authorship, and sequences of change can be inspected without assuming that every question has been resolved.
Open imaging is particularly important because the altarpiece remains a religious object, a major artwork, and a civic symbol. Its stakeholders do not share one priority. Worship, tourism, art history, public education, security, and long-term preservation place different demands on display. The transparent record cannot reconcile those demands, but it makes the grounds for intervention more accountable.
The project also renewed study of Hubert van Eyck. KIK-IRPA reports that technical and inscription research supports the authenticity of the quatrain crediting Hubert with beginning the work. Precisely which passages are his remains less certain. A conservation discovery can reopen attribution without assigning each brushstroke.
What this case supports
The Ghent campaign establishes that substantial later overpaint covered preserved original material and that a carefully tested treatment could remove much of it. Multi-modal imaging, sample analysis, microscopy, and staged cleaning converged on the layer sequence. The recovered Lamb's face is among the clearest visible results, not a freestanding curiosity.
The case does not supply a universal rule that earlier paint should take priority. Some overpaint protects a damaged layer; some later alterations have acquired independent significance; some originals cannot be safely exposed. Decisions depend on condition, documentation, ownership, function, and the stated aims of treatment.
Nor has technical study answered every authorship question or reconstructed the panels' appearance in 1432 with perfect certainty. Pigments age, losses remain, and the altarpiece's history cannot be reversed. Conservation here produced a better-supported view of one stage in the work's life. Its lasting contribution may be the combination of that intervention with a record detailed enough for others to examine.
Sources
5 sources, every one with a link you can follow.
- KIK-IRPA, The Ghent Altarpiece, the institute's technical publication on research and treatment.
- KIK-IRPA, Closer to Van Eyck relaunch, describing the imaging record and recovered Lamb.
- Closer to Van Eyck, the project's high-resolution visual and technical archive.
- Museum of Fine Arts Ghent, Restoration of the Ghent Altarpiece, on phases, partners, and public treatment.
- The conservation team, treatment synthesis, including layer evidence and the Lamb intervention.
