Technical Information

Items that are included in the Digital Maps virtual collection have been through a multi-point workflow. The first step in making maps available online is cataloging. The map cataloger examines each item and creates a MARC format record that is searchable in this virtual collection as well as in HOLLIS.

After the map is cataloged it is examined by preservation specialists from the Weissman Preservation Center of the Harvard University Library. The map is repaired and cleaned as needed, either by the WPC specialists or the paper conservator in the Harvard Map Collection.

Once the item has been treated, it can be photographed. All photography for the Digital Maps virtual collection is done by the Imaging Services group. The paper maps are scanned at a copy stand on a high-resolution scanner (Betterlight, Super8K2 scanning back camera and ViewFinder 5.x capture software). A vacuum easel is used to keep the maps flat during photography. The imaging specification was designed to produce detailed "Archival Master" images that, to the extent possible, are faithful reproductions of the originals, and that allow for highly detailed screen reproduction and print reproduction at up to 1:1. When possible files were batch processed using an "action" script derived from edits applied to match a representative sample. Color and tonal corrections were made using Adobe Photoshop. Image files were viewed on a calibrated monitor and editing was performed in an ISO 3664 compliant proofing environment. Master Image files were corrected and archived as TIFF files in the sRGB IEC611966-2.1 color-space

After photography, some maps are georeferenced, that is, encoded with real-world coordinates that allow the map to be displayed in proper geographic space. If the map includes coordinates in latitude-longitude, or some other coordinate system, those coordinates are used to georeference the map. Otherwise, coordinates for identifiable features on the map are obtained from modern sources and georeferencing is based on those coordinates. A minimum of four points with known coordinates is needed to georeference a map, however more points are routinely used, and the maps in this collection are generally georeferenced to six to ten known locations. An affine transformation is applied, which corrects for scale, rotation, and location. No rubbersheeting or warping of the images is done during this process. For more information on georeferencing historic maps, see the Harvard Map Collection's powerpoint on the topic. Some maps are remarkably accurate and will align with modern information very well. Others are less accurate and don't quite fit with the lay of the land. The main goal of the georeferencing done for the Harvard Geospatial Library is a rough fit that will facilitate geographic searching, so somewhat inaccurate maps are georeferenced and included in the collection.


Page Last Reviewed: July 12, 2017