Images and Metadata
The images provided in the database are in 300dpi JPEG format. We have 600dpi TIFF images of both adjusted and unadjusted (see the workflow above) scans for archival purposes; if you are interested in such files, please contact us at
Some brief notes on the metadata accompanying the entries will be useful. Subjects ranging from the general (“Decree”) to the specific (“Decree, Citizenship”) are included for each inscription, and a short description of the contents of the inscription is given in cases where any specifics can be gleaned. Users can browse or search by these subjects. Entries also include a general date (where known), giving the century or centuries in which the inscription is thought to have been produced. The format for this date is the century in Roman numerals, followed by an “a” for dates BCE and a “p” for dates CE; e.g., IVa would refer to the 4th century BCE, while IIIp would refer to the 3rd century CE. If a more specific date is known or has been theorized for the inscription, this specific date is also included in the form given in IG II(2). The location and language(s) of the inscription are also given, as are links to the Packard Humanities Institute's tool, which includes a published text of the inscription, and to which includes an English translation and some notes (but does not exist for all, or even most, inscriptions).. We can also carry out (RTI) on small (~10cm x 6cm) subject areas in an effort to improve the visibility of minute three-dimensional features. Again, if you think that such an exercise would be of value to your work, and we will work out arrangements.
In cases where inscriptions have been associated or joined to each other subsequent to the publication of IG II(2), including newly discovered fragments, we have employed the following procedure. If an inscription has its own IG or SEG number, it is given its own entry in the Krateros database. This entry includes links to any other entries that have been associated with it, and will cite the publication of the association in the entry’s “Description” field. Its metadata will otherwise be identical to the inscription with which it has been joined/associated. If an inscription (e.g., from the Agora or the Epigraphical Museum) has been published as joining or associated with an IG or SEG inscription without receiving an IG or SEG number of its own, the images of any squeezes of that inscription are simply included in the entry for the IG or SEG inscription with which it has been joined/associated, with publication of the association again cited in the “Description”.
Please note: although conventional practice is to indicate the edition of a published work with a superscript Arabic numeral (so, e.g., the third edition of Inscriptiones Graecae volume I would be IG I3), on this website and in the Krateros database we instead place the Arabic numeral in parentheses. Thus, the third edition of Inscriptiones Graecae volume I is rendered IG I(3). We have elected to take this approach for the sake of consistency, as the metadata fields in our database do not permit the use of superscript formatting.
The images of the squeezes are available for use, but please indicate their origin (including the name of the squeeze collection and the Institute for Advanced Study) and their Krateros ID in your publication. Please also send an electronic copy of your publication to .
The squeezes are scanned in greyscale on a , utilizing its 3D-Lighting feature. In all cases they are scanned twice: once vertically and once after a 90° rotation. The reasons for this are twofold. First, the 3D-Lighting is adept at revealing geometries that are perpendicular to the lightsource, but less so for geometries that are parallel to the lightsource. By providing two rotations, some surface features that may not appear in one image may appear clearly in the other. Second, the two images can be used with the created by The Digital Epigraphy and Archaeology Group at the University of Florida to produce a full three-dimensional representation of the squeeze.
The squeezes are subsequently subjected to some minor adjustments in Adobe Photoshop: (1) they are rotated so that their text faces right-side-up; (2) they are unmirrored, so that their text reads in the same direction as the original inscription; (3) their brightness is adjusted down and their contrast up by a uniform amount to improve readability. In cases where a squeeze is larger than the scanner bed, multiple scans are taken and then "Photomerged" together in Photoshop to form a single image, which is then adjusted as usual.
As mentioned above, we digitize squeezes with a , utilizing its 3D-Lighting feature. Our process of scanning all squeezes twice, once vertically and once after a 90° rotation, allows the possibility of using the created by The Digital Epigraphy and Archaeology Group at the University of Florida to produce a full three-dimensional representation of the squeeze. At present, the Krateros Project itself does not process its images through the Digital Epigraphy Toolbox or display the resulting 3D models in our database. We may begin doing so at some point in the future, but for the time being scholars interested in the added visualization benefits provided by 3D reconstruction are encouraged to download our images and test out the Digital Epigraphy Toolbox for themselves.
In addition to scanning, we have explored other imaging technologies. Photography was considered, especially for the larger squeezes, but presented difficulty in achieving the desired resolution. Reflectance Transformation Imaging was also investigated, but although the resulting images were promising it was difficult to craft an efficient process for applying the technology to a collection as large and varied in size as our squeeze collection. Nevertheless, we are able to carry out (RTI) on small (~10cm x 6cm) subject areas in an effort to improve the visibility of minute three-dimensional features, and if you think that such an exercise would be of value to your work, . Photogrammetry was attempted, but was found to provide poor visualization of the surface geometries of the squeezes.
If you know of another technology that you think would be of use in imaging squeezes, please send us an email and we will look into it.