Buried Texts: Multispectral Imaging at Göttingen

In the late 1990s, a rare manuscript fetched a remarkable $2 million at auction. It was a codex, dating from the mid-tenth century, and parts of it were forgeries – someone had added four portraits of the Evangelists onto its pages in the early 20th century, presumably as an attempt to increase its value. But the book was valuable not for these illuminations, or even the religious text which covered most of its pages. It had become one of the most expensive manuscripts of all time because of the text that had been deliberately erased from it in the thirteenth century: a collection of works by Archimedes, including the only extant text of his Method of Mechanical Theorems.

Erased and overwritten, damaged by water and mold, covered up by forged illuminations, not all of this text was illegible. The manuscript’s anonymous American buyer, in partnership with the Walters Art Museum in Baltimore, thus convened a group of imaging specialists to recover what couldn’t be read. The result was the first modern multispectral imaging program, and the discovery of several texts lost to time.

Multispectral imaging (MSI) is, at its most basic, simply the capture of image data at specific points along the electromagnetic spectrum, usually including the ultraviolet and infrared regions. UV and IR will be familiar to any manuscript scholar, and have been long deployed to recover illegible text from manuscripts and cultural heritage objects. But those technologies have limitations, and often fail to fully recover lost text. Modern MSI systems capture light at a number of discrete wavelengths, or bands, from the ultraviolet through the visible and into the infrared. Contemporary systems use LED illuminations and short exposure times, limiting the amount of time any given folio of a manuscript needs to be exposed to light and preventing any significant heating of the material. 

Particularly valuable for the study of palimpsests is the use of a transmissive light source, which directs light upwards through the page of a manuscript. Iron gall ink, the most widely used ink in medieval Europe, is slightly corrosive, and will eat into parchment over time. Even if all chemical traces of the ink have been removed, then, the folio will still be slightly thinner in the places where the ink once sat. Transmissive light can reveal these ‘ghost letters’. 

Filters may also be used to refine these bands still further: for example, one may shine UV light on a manuscript and deploy a filter that blocks UV light entirely, so that the only captured by the camera is that which the manuscript emits itself in response to the shortwave light. This is known as fluorescence, and is what makes your white shirt glow under a black light. 

But capturing the images themselves is only the first step. Only rarely will any one band be significantly better in terms of legibility than what can be obtained through traditional UV or IR imaging. The second phase of an MSI project is the use of statistical processing. Using programs such as Harris Geospatial’s ENVI or the freely available Hoku, the image data can be crunched and divided, refined and parsed to bring out target elements — like the erased text of a palimpsest. One such technique is principal component analysis (PCA), which is a mathematical way of isolating patterns from complex data. In 2009, PCA was used to discover another new text in the Archimedes Palimpsest, a commentary on Aristotle’s Categories.          

Finally, the processed images may be fed into PhotoShop or similar software in order to enhance contrast, adjust brightness or perform similar global tweaks in order to maximize legibility. SUB Göttingen uses a Phase One Rainbow MSI system and will soon be aquiring a transmissive light from MegaVision Inc. Though MSI has perhaps been most famously applied to medieval manuscripts, it can also be used to recover watermarks, signatures, and ownership marks from printed books, as well as identify features in artwork. 

Introductory Literature



Davies, Helen, and Alexander J. Zawacki. Making Light Work: Manuscripts and Multispectral Imaging, in: Journal of the Early Book Society 22 (2020): 179–196.

Dorado-Munoz, Leidy, David Messinger, and Damien Bove. Integrating Spatial and Spectral Information for Enhancing Spatial Features in the Gough Map of Great Britain, in: Journal of Cultural Heritage Research 34 (2018): 159–165.

Kleynhans, Tania et al. Low-Cost Multispectral System Design for Pigment Analysis in Works of Art, in: Sensors 21 (2021): 1–15.

Diesen Blogbeitrag zitieren
Alexander Zawacki (2023, 17. Oktober). Buried Texts: Multispectral Imaging at Göttingen. Selten und Bemerkenswert . Abgerufen am 19. Mai 2024, von https://doi.org/10.58079/ulkv

Alexander Zawacki

Alexander J. Zawacki is an imaging scientist and lecturer in the digital humanities at the University of Göttingen. He teaches in the Digital Paleography Spring and Summer Schools with the Institute for Digital Humanitities. His research focuses on the use of multispectral imaging to recover damaged cultural heritage documents and on magic and the supernatural in the Middle Ages. He can be reached at alexander.zawacki (at) sub.uni-goettingen.de

Das könnte dich auch interessieren …

Schreibe einen Kommentar

Deine E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert

Diese Website verwendet Akismet, um Spam zu reduzieren. Erfahre mehr darüber, wie deine Kommentardaten verarbeitet werden.

Suche in OpenEdition Search

Sie werden weitergeleitet zur OpenEdition Search