----------------------------Original message----------------------------
 
               State University of New York at Stony Brook
                       Stony Brook, NY 11794-3331
 
                                            David Y Allen
                                            Library-Reference
                                            516 632-7110
                                            12-Jul-1995 09:52am EDT
FROM:  DYALLEN
TO:    Remote Addressee                     ( [log in to unmask] )
TO:    Remote Addressee                     ( [log in to unmask] )
 
Subject: Re: Priorities for Digitizing Maps
 
LONG MESSAGE.  PLEASE EXCUSE THE CROSS-POSTING.
 
        Alice Hudson asked for people's ideas concerning priorities for
digitizing maps.  The subject is quite complex, as I am sure Alice knows (in
spite of her professions of being a neophyte in this area).  Here are some of
my thoughts on the subject, which I hope will provoke further comment and
discussion.
 
        First, the question of priorities cannot be addressed without first
examining several other questions.  These include:  What is the purpose of
digitization?  What resolution and standards will be used?  How will the
digital images be distributed (Internet, CD-ROM, printed images)?  Is the
project under consideration going to be a small one or a large one?  And, not
least important, what are other libraries doing?
 
        At present there are no standards for digitization.  Most map images of
a size convenient to transmit over the Internet are not adequate for research
purposes, although they are useful for identifying maps and getting some idea
of what they look like.  If this is the purpose, it has major implications for
the number of maps to be digitized and the method use to digitize them.
 
        If the image is to be used for research (i.e. if there is a need to
identify such things as contour lines and place names), it will have to be
distributed on CD-ROM.  The method of digitization will depend, among other
things, on the size of the map.  Regular Kodak Photo CD, for example, does not
appear to be adequate for producing "research level" images of maps larger
than, at most, two by three feet.
 
        My own feeling is that any serious digitization project of historic
maps should aim at producing images at a variety of resolutions, such as is
done by Kodak Photo CD and Kodak Pro Photo CD (although these are not the only
possible options).  It is important to pay attention to issues such as color
balance by including a color bar and a gray scale.  The most important work on
these issues to date seems to have been done by the Columbia oversized image
scanning project.  (For more info:
http://www.columbia.edu/imaging/html/largemaps/oversized).  Unfortunately,
there are no accepted standards as to what is an acceptable digital image for
research purpose, and I would propose that an effort be made to develop them.
 
        To finally get to the heart of Alice's question, it is important to
know what is being done elsewhere.  As some of us learned at ALA, the Library
of Congress is undertaking a major project to digitize some of its historic
maps.  This project may lead to the development of some de facto standards.
More important, there is no sense in digitizing something that is about to be
done by LC or that is an obvious candidate for a cooperative project by major
institutions.  It does not make a lot of sense to me for individual libraries
to go about digitizing portions of very large and well known collections, such
as Serial Set maps, early U.S.G.S. topos or geological maps, or Sanborn maps.
 
        Thus, I would suggest that individual libraries concentrate on maps of
their state or region which are not included in the major sets.  (How about
maps of NYC Alice?).  A number of libraries have already undertaken such
projects.  I know of projects at the U. of Georgia
(http://scarlett.libs.uga.edu) and at the U. of Arizona
(http://www.library.arizona.edu/images).  There is also my own project on
pre-1830 maps of New York State (a detailed description of this will be
forthcoming withing the next couple of months).  There may well be others.
Here, too, it is important to avoid duplicating work, and it is unfortunate
that there is no centralized listing of these projects.  There is a category
for "cartographic images" on the Image Clearinghouse at the U. of Arizona Web
site mentioned above, but there is still only one project (my own) listed on
it.
 
        This may not be the best time to launch a lot of new projects.  This
appears to be an area that is about to "take off," and the situation may be
much clearer in a year or two when more results are in from the Library of
Congress project, the Columbia project, Project Alexandria at Santa Barbara
(http://alexandria.sdc.ucsb.edu, soon to change to alexandria.ucsb.edu), and
others.  We are still in the experimental stage, and a lot of work may have to
be redone if it does not meet adequate standards for producing "research level"
images.