AIC's 46th Annual Meeting

With the conservation treatment of the two pendant portraits by the Dutch 17th-century Master, Rembrandt van Rijn (Portrait of Marten Soolmans and Portrait of Oopjen Coppit, 1634, canvas, 207.5 x 132 cm, SK-A-5033, SK-C-1768), newly acquired by the Dutch and French Governments, the Rijksmuseum saw an opportunity to push their photographic imaging capabilities even further. Although the Rijksmuseum has a well-established imaging protocol, including consistent lighting and end-to-end color management, it was decided, given the importance of the paintings and their conservation treatment, to utilize high resolution (1200 ppi) and multiple imaging modalities. Additionally, in order to better understand the physical condition of the pictures, they were imaged at each stage of conservation using the same imaging techniques, facilitating high-precision stitching and registration of images of both paintings and across wavelengths. The stitched and registered images, each exceeding 6 gigapixels, were then visualized using the “curtain viewer”, an internet-based image viewing technology developed by Erdmann for the Bosch research and conservation project. For this high resolution photography the museum faced several challenges. For the overall images of these large paintings, a total of 242 images were required. This amounts to approximately 70 Gb per painting per imaging modality. The maximum storage capacity of the Rijksmuseum’s Digital Asset Management (DAM) software is currently only 2 Gb per file. It was also an enormous undertaking to stitch such a large number of composite images and register them for use in the curtain viewers, enabling the conservators to fluidly switch from an overall image to the micro level and back using only the mouse wheel. In this way different technical and chemical images of the paintings, including X-radiographs, ultraviolet fluorescence images, infrared photographs and reflectograms and elemental maps acquired with macro-XRF scanning could be selected and compared ‘side by side’ using the curtain viewer. Despite the challenges, the images have been indispensable for the conservation treatment, providing insight into painting technique and condition including degradation phenomena. For example, lead soap aggregates can clearly be discerned in the high resolution visible light images, and remnants of older coatings in the high resolution UV images. By comparing photographs from before, during and after treatment the conservators could precisely track the area of interest during different stages of treatment. This paper presents the impressive benefits for conservation and research of such a large multimodal data set resulting from the combination of high resolution and multiple imaging modalities. We argue that such imaging strategies could serve as a standard in the future, both for art-historical and conservation research, as well as for comprehensive documentation of the state of a painting and its treatment.