Smithsonian Collections Blog

Highlighting the hidden treasures from over 2 million collections

Collections Search Center

Friday, December 13, 2013

Upgrading Image Links

The Archives Center has created thousands of catalog records in SIRIS, many with linked image files.  Unfortunately, many of those images are of embarrassingly low quality.  Some of you may have run across examples while browsing the database and wondered why they look so ghastly.  Now, at last, the truth will be revealed!

Recently scanned image from the Underwood & Underwood Glass Stereograph Collection.
Negative, half of a stereo pair, photographer unidentified
The problematic images are from the Underwood & Underwood Glass Stereograph Collection, which contains primarily original stereoscopic negatives and interpositives.  Because these glass plates are fragile and difficult to view, they were good candidates for one of the Archives Center's first forays into the creation of a visual surrogate--a videodisc or laserdisc.  Remember videodiscs?  They were once considered cutting-edge technology--after all, they employed lasers--and the Smithsonian's Division of Photographic Services and the National Air and Space Museum were enthused about their potential as research aids.  Commercial motion pictures were issued in videodisc form, and my cousin has shelves filled with hundreds of operas and concerts on videodisc; I certainly hope his videodisc player is still functioning.  (The Archives Center's laserdisc players died some time ago, with little hope for replacement.)  The videodisc was actually quite versatile, as 54,000 still images could be encoded on each side for a grand total of 108,000.

The first Archives Center videodisc contained color transparencies from the Donald Sultner-Welles Collection, and the surrogate was intended to provide rapid access to the photographs while the originals were in cold storage.  There was no database, but videodisc frame number sequences were recorded laboriously for folder-level descriptions in the 350-page finding aid.  Armed with these frame numbers, one could zip swiftly to specific images and image groups, or one could browse the videodisc at random, then look up descriptions by frame number in the finding aid.

The Underwood & Underwood collection was accompanied by a database, however.  The Museum's IT staff persuaded us that this was the way to go, and it represented a technological and philosophical breakthrough for us.  Using dBaseIV, we catalogued all 28,000 items in an incredibly short time.   Thanks to a grant, we hired a well-known contractor in the field to copy all the plates onto 35mm film with a custom-made animation camera; the film was then transferred to videotape, from which a videodisc was created.  My original assumption was that access to the videodisc images and accompanying descriptive data would be accomplished in the Archives Center on a stand-alone computer connected to a videodisc player (possibly forever), and our collection-level SIRIS record stated that procedure.

When we were later able to export our dBaseIV data into SIRIS, online access to these images seemed feasible at last.  We set about digitizing the Underwood & Underwood plates from the videodisc, using an inexpensive screen-capture device called "Snappy."  The procedure was surprisingly quick, but it resulted in surprisingly low-quality surrogates as well.  Think of how many generations these images represented: they are digital files created from analog video images from a videodisc, which in turn had been created from a videotape copy of a 35mm film copy!  (Never mind the inherently low resolution of "Snappy" files!)

"Medium-ghastly" image from SIRIS.  "How Biddy served the tomatoes 'undressed.'"
Photograph is a stereoscopic negative from the Underwood & Underwood Glass Stereograph Collection,
originally published by H.C. White & Co., ca. 1910.

The vast majority of these poor-quality Underwood & Underwood images remain online, although some have been replaced by new scans from the original plates on an ad hoc basis as researchers order copies, and the difference is enormous.  These replacements seem to proceed at a glacial pace, considering the high number of poor copies remaining online.  These old images remain an embarrassment and viewers have complained, but we feel that they're better than nothing.  What do you think?

David Haberstich
Curator of Photography, Archives Center

1 comment:

  1. how about a systematic effort to digitize images; upload to flickr or wikimedia commons.
    volunteers can be marshalled, like the http://www.si.edu/Volunteer/DigitalVolunteers

    ReplyDelete