The Two Faces of “Digital Emblematica”
Nuala Koetter
Digital Services and Development Unit
University of Illinois Library at Urbana-Champaign
The University of Illinois Library owns an extensive and impressive collection of Emblem books. Cataloged in 1993 by Thomas McGeary and Fred Nash[1], the collection then held more than 600 original editions and 400 emblem books in reprint or microform format. Since 1993, the collection of original emblem books has grown to over 650.
As an access tool, the goal of the “Digital Emblematica” project is to develop a searchable image database intended for an international audience of emblem scholars and students. As a preservation tool, the web site facilitates preservation of the emblem books by reducing the necessity to see the actual book. In this paper, we will discuss the method by which the books are being digitized and their images included on the web site, and secondly, the metadata schema and database structure as well as the publicly accessible search interface.
Digitization
The
University of Illinois Digital Emblematica project was originally started in
1998 when one emblem book by Johann Pfann was digitized. In the late 1990’s, we scanned this book
using a Hewlett Packard ScanJet 4C Scanner (Resolution 600 / 2400 dpi; Pixel
Depth: 30-bit / 10-bit internal). The
software used at the time was HP DeskScan II, managed with Visioneer PaperPort. Digital photographic images were produced
using a Leaf Digital Camera Back (II) and a Mamiya 2Z67 pd Prismfinder camera
front and lens. We also scanned a
number of sample emblems from other books.
Maximum image file capture was 16 MB, and resolution was captured from
200 to 450 dpi, depending on the size of the original object. Using a three pass RGB scan, pixel depth
could be captured up to 30 bits.
The static images from the Pfann book were then made available on the project web site at http://images.library.uiuc.edu/projects/emblems/[2]. In 2002, due to receipt of new project funding (described by Kilton[3] and Billings[4]), we were able to continue digitization of the Library’s emblem book collection. Due to more updated equipment and facilities, we are now able to digitize more emblem books from our German emblem collection and also generate better quality images.
Images are now being captured either by digital camera or on a flatbed scanner, depending on the condition of the book. For those books that are in a reasonably good condition, images are captured using a Nikon D-1 digital camera[5] and Nikon Capture 2.0.0 imaging software. Master images are captured in TIFF format, 24-bit color, using a 28-70 mm Nikkor lens. Image size on average is 2000x1312 pixels (roughly equivalent to 300 DPI). The master image size averages 6.0 MB. Images are then optimized using Adobe Photoshop software[6]. For this project, image modification is actually kept to a minimum. Images are rotated, cropped, and light levels are adjusted to reflect the look of the original piece. Access images are then created as JPEGs with minimum compression and sized to 1024x768 pixels.
For those books that are more robust, images are captured using an Epson Expression 1640XL flatbed scanner[7] and Silverfast scanning software. In this case, master images are captured in TIFF format, 24-bit color at 600 DPI. The final master image size often ranges from 25 to 120 MB. Again, images are optimized using Adobe Photoshop software and image modification is kept to a minimum. Where necessary, images are rotated, cropped, and light levels are adjusted to reflect the look of the original piece. As with the digital photographs, access images are created as JPEGs with minimum compression and they are sized to 1024x768 pixels.
Metadata Development
Once the images have been created, the TIFF images are archived and the JPEG images are uploaded to the online database. The online database, which we are using for the emblem collection, is a commercially available product, CONTENTdm[8]. Originally developed at the University of Washington, Seattle, this image database software has been made commercially available through DiMeMa, Inc. The CONTENTdm software simplifies the process of digitizing both graphics and text materials by providing an easy-to-configure and manage database environment for acquisition, indexing, and the design of Web-based search access to digital collections. The software is also XML-compliant, which will provide for reasonably easy transport of the data from CONTENTdm to other applications in the future.
The software allows for standard metadata fields based on the Dublin Core[9] best practice, advanced searching capabilities, and data entry from remote locations. This makes it a very suitable package for the digitized emblem book collection. As well as being the default format that comes with the CONTENTdm software, the fifteen field Dublin Core[10] metadata schema was chosen for a number of reasons. In particular, it is simple to use for those who are unfamiliar with the often difficult concept of metadata. Dublin Core uses commonly understood terminology. It is also international in scope and, more importantly, it is extensible and can be manipulated according to the needs of a particular project.
In the case of the emblem book project, we felt that several of the Dublin Core field names were not fully descriptive of the information for which they would be used, so we renamed a number of fields more appropriately. In addition, Dublin Core allows the repetition of any field any number of times, as shown in Table 1. For example, the Motto field, for the purposes of this project, can be in Latin, German or Early Modern German. Each of those fields is treated separately but they are each mapped back to the Dublin Core Description field.
For the emblem book project, we have developed metadata based on individual emblems and on the whole emblem book. One drawback of the CONTENTdm software is that it does not allow us to generate two separate tables, one for the book and one for the individual emblems. Necessarily then, some fields may at times be empty as they are not being used for a particular record. For example, where we are describing individual emblems, it may not be appropriate to give the Book Memo information every time.
Table 1 describes in detail the metadata fields that are currently being used in the database. The first column, Emblematica Field, describes the title of each field that is used in the database. The second column, Dublin Core mapping, shows the Dublin Core field to which the emblem field is mapped. Following that is the description of the field in the third column. The fourth column shows if a particular field is searchable by a user or not.
|
Emblematica Field |
Dublin Core mapping |
Field Description |
Search? |
|
Title / Emblem ID |
DC.Identifier |
This is the field that
corresponds to the image file name.
If the record is an emblem, an identifying number is used to associate
the image to the record. Similarly,
if a record is an emblem monograph, an identifying number or name is entered
to correspond to the monograph title. |
Y |
|
Short title |
DC.Title |
This is the given
transcription of the most important elements of the title of the
monograph. It may also include
information about the edition. |
Y |
|
Record No. |
DC.Identifier |
Emblem record number |
N |
|
Motto (Latin) |
DC.Description |
The Motto is given in three
forms where necessary, Latin, Early Modern German and German. In this first Motto field, the Latin Motto
is transcribed as in the source. It
is transcribed using the old spelling and diacritical marks, although
abbreviations are expanded as necessary. |
Y |
|
Motto (Early Modern German) |
DC.Description |
The Early Modern German
Motto is transcribed appropriately with syntax. |
Y |
|
Motto (German) |
DC.Description |
The Motto is repeated but
with the spelling modernized according to modern German spelling. |
Y |
|
Descriptors |
DC.Subject |
The relevant objects,
activities, persons and personifications allegorical figures, topoi, etc.
appearing in the emblem are listed in this field. English is used where possible, but common German or Latin
terms for topoi may also be used. |
Y |
|
Topos |
DC.Subject |
Topos includes composite images. |
Y |
|
Theme |
DC.Subject |
The Theme field is used to
describe the abstract moralistic themes deriving from the images. |
Y |
|
Artist |
DC.Creator |
The Artist field is used to
hold information about the artist, designer and engraver/etcher/wood carver
of the emblem. |
N |
|
Book Identifier |
DC.Identifier |
In this field, the least
ambiguous means of locating the complete emblem (image and explanatory text)
in the original book is given. This
might include the emblem number and/or inclusive pages, as appropriate. |
N |
|
Author |
DC.Creator |
The common form of the commonly
used author, editor, compiler or publisher of the book. |
Y |
|
Place of Publication |
DC.Publisher |
The English form of the
place of publication is given. |
N |
|
Publisher |
DC.Publisher |
The common names of the
publisher and/or printer are given, including all names used in the imprint. |
N |
|
Date of Publication |
DC.Date Created |
For a multi-volume book,
inclusive dates or dates of each volume (as appropriate) are given. |
N |
|
Bibliographic Identifier |
DC.Identifier |
Where possible, the numbers
in emblem reference books of the University of Illinois are included, e.g. McGeary, T.; Nash, N.F.
Emblem books at the University of Illinois: a bibliographic catalogue.
Boston, Mass.: G.K. Hall. 1993. |
N |
|
Book Memo |
DC.Description |
This field is used for any
comments relating to the entire book’s author, provenance, bibliographic
problems, copies consulted, etc. |
N |
|
Technical Details |
DC.Description |
Within this field, the details of the digitization process for the book are described. In particular, the type of digitization process and the resolution of the image are noted, together with any details of image manipulation. |
N |
|
Rights |
DC.Rights |
Rights management statement
appropriate to the material. |
N |
|
Location |
DC.Publisher |
Publisher or owner of the
material. |
N |
Table 1.
Metadata and field descriptors for emblem books
Once the
metadata for each individual emblem has been generated, we upload the images
and corresponding metadata into the CONTENTdm image database through an
Acquisitions Module. The individual
emblems are then brought together and generated into one monograph record,
availing of the software’s page-turning abilities.
Searching for Emblem Records
Once a record
has been uploaded, users can immediately search for particular emblems or
emblem books. Through a browse facility
on the search engine, they can also browse through the available titles. Figure 1 shows the user interface[11]
to the database with the drop-down facility.
This drop-down menu allows the user to see a listing of all the records
and information that are contained in a particular field.
Fig. 1:
User search CONTENTdm interface to emblem database, showing the
drop-down list of searchable fields
For example,
if the user should choose the Title/Emblem ID field, he or she can see a list
of those emblem books that are already digitized and publicly accessible. By clicking on one of those listed books,
the user is then shown a thumbnail image for the emblem book. Subsequently clicking on the thumbnail image
will bring the user to the page-turning facility on the image database, as
shown in Figure 2.
Fig.
2: Emblem monograph with links in the left frame to the individual
emblems. The description of each
individual emblem is linked below the image.
The left frame in Figure 2 has links to each of the individual emblems that are available for a particular monograph. Below the list of links is a link to the description of the emblem book. In the right frame, the actual emblem images can be seen and they will obviously change as the user clicks on the links in the left frame. The description of each individual emblem is then linked below the emblem image.
Figure 3a shows an individual emblem and Figure 3b shows the corresponding metadata for that emblem. The descriptive information about the emblem image is entered to the appropriate fields for that image. Under the Descriptors field, there is a list of the individual descriptors used to describe the image. In this particular example, it was not appropriate to include a Latin Motto, so that field remains blank.
Fig. 3a:
Individual emblem from “Emblematica Politica”. Descriptive information for this emblem can be seen in Figure 3b.
Fig. 3b: Descriptive information, based on
Dublin Core format, for the emblem in Fig. 3a.
The image software that we are using for the “Digital Emblematica” project also allows the user to compare and contrast two images. In order to do this, the user selects two images through the “My Favorites” link in the top frame of the search page, and puts them side-by-side on the screen, allowing comparison in one open window. Figure 4 illustrates how this can be done.
Fig. 4: Example of a number of images, which
were chosen through the “My Favorites” link[12]. Two of these images can then be compared
with one another, or all favorites can be chosen to generate a new Web page for
the user.
In addition, the software allows the user to mark as many images as desired, from one or more books, into a “My Favorites” link. From this Favorites page, as seen in Figure 4, the user can then generate his or her own Web page with those images and use that page at a later date for research.
Future Work
As we develop the Digital Emblematica collection, a number of issues are arising which we are now considering. Firstly, we plan to start including the Iconclass classification system. Iconclass is an international classification system specifically designed for “iconographic research and the documentation of images”[13]. Including the Iconclass feature will allow us to be a lot more descriptive of the emblem images and to ensure that we are describing the images according to best practices carried out by other institutions.
In addition, we are currently in conversation with the Herzog August Bibliothek in Germany[14], discussing our mutual interests in developing multi-lingual vocabularies, as well as our use of the Dublin Core metadata schema and their use of TEI, the Text Encoding Initiative, schema.
Finally, we are considering expanding this project in conjunction with the Open Archives Initiative project at the University of Illinois Library[15]. This Open Archives Initiative project, which has to date been funded by the Mellon Foundation, is creating and implementing a suite of metadata harvesting services, search services and tools designed to facilitate discovery and retrieval of scholarly works, in particular, cultural heritage materials.
It is appropriate that we consider the Digital Emblematica project in this context and allow the harvesting of the emblem metadata, thereby attracting an even wider audience than we currently have.
Acknowledgements
Funding for this project was provided by the University of Illinois Office of the Vice Chancellor for Research. I would like to thank Beth Sandore, Mara Wade, Tom Kilton, Marshall Billings, Trevor Jones, Gene Rinkel and Madeline Gibson for their valuable assistance in making this work possible.