, , , , , , , , , , , , ,


The Martellus map on an easel, which allowed the object to be repositioned for the camera. Image courtesy of Chet Van Duzer.

In 1962, thanks to an anonymous donation, the Beinecke Rare Book and Manuscript Library at Yale University acquired a 15th century world map created around 1491 by German cartographer Henricus Martellus. According to Michael Phelps, director of the Early Manuscripts Electronic Library in Rolling Hills Estates, California, the map – or a version of it created by Martellus – is likely one that Christopher Columbus consulted prior to embarking on the 1492 voyage that landed him on the shores of the New World. A map made in 1507 by another German cartographer, Martin Waldseemüller, makes use of the Martellus map, Phelps said, explaining that the latter represents “a real turning point in history.”

The problem with learning more from the Martellus map, until this point, had been that to the naked eye, Phelps said, “it’s indistinct.” Time has caused the map to fade. Phelps said that in the 15th century, rivers and cities on the map would have been in great contrast, in terms of the colors that were used to depict them, but that over time, those colors have become muted and (again) indistinct. Until now.

In August, a team of researchers led by independent scholar Chet Van Duzer spent 10 days at the Beinecke using a multispectral imaging process to bring out the history of the map. Mike Cummings, the Beinecke’s public-relations manager, said the library welcomed the multispectral imaging team. “We support all kinds of scholarship and we’re excited to see what they are able to produce,” Cummings said.

Van Duzer said he first began studying the Martellus map as a source for the 1507 Waldseemüller map that resides at the Library of Congress. The significance of the Waldseemüller map, Van Duzer said, is that it was the “first map to apply the name America to the New World.” Wanting to compare the two maps, Van Duzer first studied ultraviolet images that were taken of the Martellus map in the early 1960s – before the map was acquired by Yale University. In 2010, after staff at Yale the previous year made new ultraviolet, infrared, and natural-light images of the map, Van Duzer studied those images at the independently run John Carter Brown Library, on the campus of Brown University, in Providence, Rhode Island. Able to read about one-third of the map’s previously illegible text, Van Duzer wrote a book on the subject.

“It’s one of the most important maps of the 15th century,” Van Duzer said. “I’ve been sitting on the book for three years now,” he said, waiting for the results of the multispectral imaging process that was conducted at the Beinecke in August. He’s in the process of expanding on his book about the Martellus map. The technology for the Martellus map multispectral imaging project was made available by Gregory Heyworth, a professor in the English department at the University of Mississippi and the director of the not-for-profit Lazarus Project, which connects researchers with portable, high-end multispectral imaging equipment. The project itself was funded through a grant from the National Endowment for the Humanities. Whereas in 2010 he spent a week in Providence trying, with the 2009 images taken at Yale, to read a block of text in the lower-right corner of the Martellus map, Van Duzer said images captured in August have allowed him to read that same text “without any difficulty whatsoever.”

Still, he said, “the processing of the images is an art” that takes a while. At press time, only a handful of the images captured at Yale had been processed. Eagerly, Cummings said he and his colleagues at the Beinecke will “be able to put the images on our website, and we’re excited for that.” Phelps, in an email, explained the multispectral imaging process in detail.

“Multi-spectral imaging,” Phelps wrote, “involves capturing a set of images of a single object at different wavelengths or bands of light, in order to better discern information about the object. Colors of light, including those the human eye can see and those it cannot, can be arranged along a spectrum according to their wavelengths. On this spectrum, the human eye perceives a broad band of different colors or wavelengths of light. … Multi-spectral imaging [MSI] involves capturing individual images at specific wavelengths or narrow bands along this spectrum, in order to discern new information about an object.”

Roger Easton, Phelps wrote, a professor of imaging science at the Rochester Institute of Technology who, along with Phelps (and Heyworth), sits on the Lazarus Project’s Board of Directors, developed the technology, which “creates derivative images that combine data from the captured images.” The technology’s “application to cultural heritage is certainly new,” Phelps explained. A most important part of this kind of project, he pointed out, is the “feedback loop between scholar and scientist” – that is, in this case, between Van Duzer and Easton – through which the team works “to try to extract as much information as we can from the information we extracted from the map.”

In his above-mentioned email, Phelps explained that Easton “processes the captured images of the Martellus map in order to generate a series of derivative, processed images which maximize the legibility of text and other information on the map. Roger builds an ‘image cube,’ in a sense a stack of the captured images, and then uses statistical methods to analyze the collected data and to distinguish features of the map. Some of these features, such as faded writing, may be illegible to (the) naked eye, but can be isolated statistically and then rendered legible in processed images. A feedback loop between Roger Easton, project scientist, and Chet Van Duzer, project scientist, improves the processed result and insures that we maximize the legibility of faded and obscured text on the map.”

Heyworth pointed out that capturing the images is only 15 percent or 20 percent of the process. The majority of a multispectral imaging project involves the image processing and the task of reading and understanding what’s revealed.

“Multispectral imaging is an esoteric science,” Heyworth said, explaining that “we do textural science,” and that “the technology … is in its inception.”

The goal, he said, is “to reveal writing that is not visible … and also to create a highly color accurate … digital reproduction of the object,” a reproduction that reaches back in time. He likened the process to looking at stars. Through the multispectral imaging process, one “looks into the earliest history of the object,” he said. And through that technology, he said, “we’re changing the canon.”

The Martellus map multispectral imaging project at the Beinecke Rare Book and Manuscript Library was a collaboration between the Early Manuscripts Electronic Library, which oversaw the imaging process, and the Lazarus Project, which provided the equipment. The imaging team included independent scholar Chet Van Duzer, who led the project; Gregory Heyworth, a professor of medieval studies at the University of Mississippi and the director of the Lazarus project, who conducted the imaging; Michael Phelps, the director of the Early Manuscripts Electronic Library, who served as the project manager; Roger Easton, a professor of imaging science at the Rochester Institute of Technology and the imaging team’s project scientist; and Kenneth Boydston, CEO of the digital imaging company MegaVision, who designed the camera used in the imaging process and helped to capture images of the Martellus map.

Learn more about the Early Manuscripts Electronic Library at emelibrary.org. Learn more about the Lazarus Project at lazarusprojectimaging.com. And visit the Beinecke Rare Book and Manuscript Library online at beinecke.library.yale.edu.

David Brensilver 

This article appears in the October issue of The Arts Paper. Read other stories from this issue online here.