The paper began by acknowledging a third author who was inadvertently left off the program and abstract booklet, but will appear as a third author in the TSG Postprints. The paper was presented by E. Keats Webb, the digital imaging specialist for the Conservation Institute at the Smithsonian. She showed how by combining a lower resolution spectral camera (Surface Optics Corp SOC710) with a modified DSLR with bandpass filters, the resulting images could be used to identify certain pigments on a painted textile. The project began with four textiles being described as plain weave with color (a limited color palette of browns and blue/blacks) delivered in a paste form without an organic binder, the color remaining on one side – this is important, keep this in the back of your thoughts as you continue, dear Reader!
I live in Silicon Valley, but I am not tech savvy. I am also several years out of school and realize that I my not have the vocabulary to give a precise recall of the methods and analysis described. Having said that, I feel I did learn about a (relatively) low cost technique that would be helpful in characterizing pigments on textiles. Essentially, the technique required several images to be taken at various bandwidths along the visible to near infra-red spectrum. Knowing the reflectance spectra for a given pigment allows the researcher to target the images for analysis. The targeted images are overlayed to get the difference which results in areas that appear to fluoresce for a positive identification.
It was found that indigo was readily identified, whether it was used alone or as a component of other colors. Reds and browns proved a bit more problematic. The success of this technique seems to have been that there was a small color palette, as well as the textiles in question were of known provenance, so they already had an idea of the pigments to look for. Also, the pigments were on one side and in good quantity. They did not achieve good results with dyes or small samples due to background noise interference. Though it was brought up in the question/answer portion that HPLC gives definitive “fingerprint” spectra for positive identification, the imaging technique presented does not require sample-taking (important to the National Museum of the American Indian) as well as allows for the entire textile to be mapped, keeping the pigment identification within context.
I could see this technique as a very useful first step in pigment identification or potentially in identifying areas of restoration and the pigments used. I look forward to reading the paper in the Postprints.
Skip to content
AIC Blog Archives: Conservators Converse
Former Blog of the American Institute for Conservation