Can imaging really read text that has been erased?

In some cases yes. Multispectral imaging records a surface across many wavelengths beyond visible light; processing can then enhance traces of ink that the eye and a normal photograph cannot distinguish, as was famously done with the Archimedes Palimpsest.

Digitization and Imaging

Turning a manuscript, painting, or artifact into pixels is a deceptively complex act. Decisions about resolution, color, lighting, and capture technique determine what the image can show — and advanced techniques such as multispectral imaging can even recover text the eye cannot see.

Definition

The process and study of capturing cultural objects as digital images, encompassing capture technologies, quality standards, and specialized imaging methods that reveal otherwise inaccessible features of the source.

Scope

Covers the technologies and standards for capturing cultural materials as digital images: scanning and photography, resolution and color management, file formats and quality standards, and advanced techniques such as multispectral and reflectance imaging used to recover damaged or erased content. Includes the relationship between imaging choices and scholarly or preservation goals.

Core questions

What capture decisions determine the scholarly usefulness of a digital image?
How do resolution, color management, and format affect fidelity and longevity?
When can imaging recover information lost to damage or erasure?
How should imaging serve both access and preservation?

Key concepts

Resolution
Color management
Multispectral imaging
Reflectance transformation imaging
File format
Capture standard

Key theories

Image quality as a scholarly variable: Terras argued that imaging is not a neutral technical step: resolution, color accuracy, and documentation determine what scholarship a digital image can support.
Multispectral recovery of lost text: Spectral imaging captures a surface at many wavelengths so that processing can separate erased or faded writing from later overtext, as demonstrated on the Archimedes Palimpsest.
Forensic materiality of the digital image: Kirschenbaum's account of digital materiality underscores that image files are themselves artifacts with formats, metadata, and histories that bear on authenticity and preservation.

History

Heritage imaging matured from flatbed scanning to high-resolution photography and computational techniques. The Archimedes Palimpsest project (1999-2008) popularized multispectral recovery of erased text, while reflectance transformation imaging spread for capturing surface detail. Standards bodies developed guidelines for sustainable, high-quality capture.

Debates

Access copies versus archival masters: Institutions balance fast, low-cost digitization for access against the higher standards needed for preservation-quality master images, given storage and budget limits.

Key figures

Melissa Terras
Roger Easton
Matthew Kirschenbaum

Seminal works

terras2008
easton2010
kirschenbaum2008

Frequently asked questions

Can imaging really read text that has been erased?: In some cases yes. Multispectral imaging records a surface across many wavelengths beyond visible light; processing can then enhance traces of ink that the eye and a normal photograph cannot distinguish, as was famously done with the Archimedes Palimpsest.