November 7-8, 2002

Site Visit: The Cornell Institute for Digital Collections

 

The trip to Cornell was very illuminating.   In terms of planning and implementation, Cornell started is projects about ten years ahead of UVM.  This gives us some advantages: we can learn from their mistakes; we have even better technology to start with; and we are better-prepared for change.

 

Interestingly, CIDC was not as centralized as may have been expected.  Most digital projects work independently of each other, and we often had the sense that people working on the projects were learning about each other's methods for the first time.  However, CIDC did provide the initial funding as well as the software (for both data and images) which is currently used by all of the projects.

 

Major funding originally came from an anonymous donor, but most projects require additional funding which is sought out by the keepers of the collections in partnership with CIDC.  

 

CIDC has strong support from the Cornell administration.

 

Trip to Ithaca —en route

 

On our trip to Ithaca , there was a very spirited discussion about the goals and visions for a University-wide digital initiative (we tended to use the shorthand "Center").   Overall, distinctions were made between the digitization process, and the uses and contextualization of the digitized object.   The Center could ease the use of digital collections by training faculty and students to use the collections within classes and research.   The creation of context and narrative collections would be facilitated by the Center.  The Center would also help unify current digitization projects within the University community.  

 

Some of the issues with organizing digital projects across campus were discussed. These included communications, and being able to learn who did what within the different departments.   Knowledge of responsibilities and projects would avoid duplication and would enable utilization of already existing resources.

 

It was noted that "Digitization" only described a part of the process.  "Digitization" was the creation of a digital surrogate, but then there was the need to make that digital object available to people who wish to use it, and to place it within a context.  Technology was described as the tool, and digitization as the process.   The center's goal could be about bringing together what had gone through the process of digitization.  Accessibility would be an important part of the Center's goal.

 

The goals for the Cornell Institute for Digital Collections were read and discussed.  These goals for the Institute included:  to create and maintain digital resources; to incorporate digital collections into the classroom, etc.; to conduct research into the utilization of digital collections; to educate and train people in the use of digital imaging technologies.   These goals seemed to fit well with what was being discussed for the University of Vermont.

 

The Center was described not as a custodian of the collections that were digitized, as much as a framework of knowledge, as had been outlined in the original proposal. 

 

The Center was described as a physical, social space where people can discuss, share ideas, and critique together.

 


 

On Thursday afternoon, the UVM group met at Kroch Library with Peter Hirtle (Director, Cornell Institute for Digital Collections), Eleanor Brown (Digital & Technology Archivist, Division of Rare & Manuscript Collections), Susette Newberry (Visual Resources Archivist), and Patrick Stevens (Curator of the Icelandic Collection).

 

Mr. Hirtle gave an overview of the origins of the Cornell Institute for Digital Collections.    He described preliminary pilot projects carried out with Kodak.  The initial organizers were Tom Hickerson and Peter Gay, who was close friends with the initial anonymous donor who has continued to support the Institute.  He described the initial impetus to be the wish to unite scattered, visually interesting materials from around the campus.  

 

The Utopia project was described.   This project was designed with faculty involvement and was carried out before any data standards had been set up.  

 

The decision to create a "generic image catalog" for the University was

  Outlined.  There were initial attempts to link this project to needs within the University administration (for example, in the tracking of correspondence).   The decision was ultimately made to use a Museum collections database produced by Willoughby called MultiMimsy 2000 for cataloguing.   This system was used by all participants (Museum, Library, Slide collections).  There were bi-weekly meetings in which standards were discussed (sometimes heatedly) by all participants.  

 

The selection of Luna Imaging to manage the University wide image collection was made because Luna permits the tools for tracking and directly working with the images.  Luna can be linked with AMICO, and also permits cross-collection searching.  

 

An example of the use of Luna by students and faculty was described.  The Herbert Johnson Museum exhibition "Friends of the Cold Season:   Pine, Bamboo and Plum in East Asian Art" done two years ago, was developed by an art history class working closely with a Museum curator and utilizing images provided via Luna. 

 

Early imaging was often done through outsourcing, but there were many issues with quality control.   Many of the digital images created by outside firms needed to be cleaned up one by one or even re-scanned.  Overall, Luna did do a better job than other consultants when it created digital images for the project. 

 

The Museum is now carrying out most of its own scanning, and has its own way to directly capture digital images from the originals [decribed later].  

 

The creation of digital images is now generally project based. For example, the Museum collections are one group; Library projects are often by exhibition or by specific, separately funded projects such as the Iceland project. 

 

CIDC funding is not on a cost-recovery basis.   Much of it comes from the single anonymous donor.   There is cost-sharing with other, project-based grants.   The Museum as well as the Library have their own departments which seek out funding.   There is also an internal, Cornell University grant program.   CIDC does assist other departments with determining digitization costs, and with writing grants.

 

CIDC has very strong support from the provost and from central administration.

 

The need to do more outreach with faculty was noted.   Some training has been done on the use of Luna InSight technology.  The use of images by freshman writing seminar classes was described as one of the best examples of the use of image resources by a class at CU.    The Slide Library and the Museum were described as the best sources for outreach to faculty.  There is a CU Academic Technology Center, but there is not much contact between this group and CIDC. 

 

Eli Brown stated that she now had a mandate for the Rare Book and Manuscript Collections:  that all exhibitions mounted had a web presence.  

 

Issues with legacy data were described.   Data standards continue to develop and technology continues to change.   The last major data conversion was to MultiMimsy .  One major lesson learned about this was:  as few applications should be utilized as possible, and file setups need to be very clear.   This permits easier conversion when necessary.

 

Possible joint efforts between UVM and CU were brought up. 

 

 

The UVM group met at the slide library with:   Margaret Webster (Director, Knight Visual Resources Facility [slide collections]), Eric Schwab (Assistant Director), Leah Theis (Curator).

 

The slide collections from the CU College of Architecture and from the Department of the History of Art have been merged over the past year.  This has brought up many issues of cataloguing, methodology, and utilization of the collections.

 

KVR was an original part of the IMAG/CU group.   MultiMimsy 2000, chosen because it met the Museum's needs, has been very problematic for a slide collection, and they will probably change over to an Access database.   MultiMimsy does allow them to share authority files and bibliographic references with the Museum.   Luna Imaging has been a good image-delivery tool since it permits campus-wide delivery and cross-collection searching.   Currently, the Luna database is not populated with enough data to meet the needs of the users of the Slide Collections.  

 

Slides are scanned in-house, and which slides are scanned is determined by faculty requests.  Slide collections by their very nature are faculty-driven.   Larger projects are outsourced.  Slides are generally scanned as TIFF files, at 18 MB.   Adjustments to the image are generally not made, since slides are considered imperfect anyway.  Files are archived on CD.

 

Images from the slide collections are used a number of ways within the classroom.  Faculty have used them within PowerPoint for presentations to their classes.  Others have used Luna for these presentations, and it is considered even better because images can be shown side-by-side and images can be placed into groups.   Image groups have been used as study aids, but are not necessarily being projected in classes.  There does need to be work to get faculty interested in the use of IMAG/CU, and set up opportunities for training.  Luna is not necessarily easy to learn; it utilizes different icons from the standard ones.

 

The James Madison University image browser was suggested for use.   This system is set up specifically for slide libraries/art history collections.  

 

Slide collections are considered teaching tools, not research tools.  Slides have now become a more university-wide resource at Cornell. There is a need to become less territorial about resources, especially as they are digitized and become part of an image database.  Now the user is not necessarily someone known to the department.

 

The slide library currently bears the costs of digitization.   Payments are sometimes made for time and materials.   It is the KVR employees who actually create basic websites for faculty and populate them with images.  This enables them to remove the websites at the end of the semester, and keep control of image use.

 

Lessons learned over the course of the project have been the need to clearly communicate.  Faculty and patrons must be clearly told and must understand the time and costs necessary to digitize images.  Time should also be allowed for appropriate cataloguing—not just for the digitization of the images. 

 

 

 

 

11/8/02

 

On Friday morning, the UVM group met with Matt Ferrari (Digital Photographer, Herbert F. Johnson Museum of Art), Rhea Garen (Digital Photographer, CIDC) at the Herbert F. Johnson Museum of Art

 

The first digital projects at the Johnson Museum involved the Print Collection of approximately 13,000 items (about 1/3 of the Museum collection).  In 1999, a 3-D camera was purchased for the documentation of the Asian collections.   The photo studio permits very strict lighting controls.   Batches of images are tracked via Excel spreadsheets.   Images are archived onto CD-R's as well as placed onto Luna.   There is very strict color control, often with the original object in front of the editor.  Each image file is archived at 70 MB, and derivatives are created at 36 MB.  

 

Priorities for photography are determined by Museum curators.   Photologs are kept manually, with curatorial needs outlined with each object (such as makers' marks).  

 

Much of the photographers' time is spent adjusting lighting and setting up each shot.  The image size is adjusted, and boundaries are determined via a live video feed from the camera to a computer monitor.  Images are captured using 30-36 tiles and 30 strobe flashes.   This permits real dimensionality and depth for the image.

 

Matt is an employee of the Museum, and he has a part time assistant who does a lot of the color adjustment.   Both are funded via project-based grant money.  

 

Museum employees use Luna to access created images, but archived images are pulled for publication needs.  

 

 

The group met with Ellen Avril (Curator of Asian Art, Herbert F. Johnson Museum), Robin Burlingham (Asian Art Cataloguer)

 

Ellen and Robin described the creation of the exhibition "Friends of the Cold Season:  Pine, Bamboo and Plum in East Asian Art."   Staff and students were trained by CIDC in the use of Luna.   Students created their own groups of images and used them to create exhibition layout and label copy. 

 

Robin creates cataloguing information directly into MultiMimsy.  This info is then linked to Luna InSight via an Access database.

 

 

The group met with Suzette Newberry, Kurt Jordan (Project Archivist, Political Americana Collection, CU Library) at Kroch Library

 

Suzette described the Library's use of MultiMimsy 2000, which permits strict authority controls and facilitates item-by-item cataloguing.  Ultimately the Library hopes to move to a Luna cataloguing module.   Currently there are about 10,000 subject terms in the MultiMimsy subject catalog. 

 

IMAG/CU has facilitated discussions between Library and Museum groups regarding issues around authorities and standards.  

 

Suzette demonstrated the Luna A.D. White architectural photograph collections.  Catalog data was taken from MultiMimsy and placed into set fields within Luna.  Currently deeper subject term searches are not possible. 

 

The web version of Luna is very stable and has few glitches.   There is an upcoming new version which will permit cataloguing directly within Luna, as well as the ability to search selected collections (rather than all of them at once).  The new Luna will also permit the use of sound files.

 

The Political Memorabilia project is funded by an IMLS grant.  There are about 5,000 items within the collection and all are being catalogued using MultiMimsy. 

 

 

Lunch

 

The Encompass program is considered to be in development and very far in the future.  CIDC became co-developers of this product (a Voyager product) a few years ago.   The anti-slavery pamphlet collection is catalogued in Encompass.  

 

Finding aids are currently being developed in HTML.   All guides are being moved to the new EAD.  

 

Tom Hickerson spoke briefly to the group.  He noted that there were far better opportunities today (technologically speaking) than there were 10 years ago. He noted the need to accommodate technological change when created digital information.  He said that there was a need for more integration within the general library system, and it would have been better if that had been done at the very beginning.  

 

 

UVM delegation:

Bill Lipke, Chris Burns, Ines Berrizbeitia , Deborah Boothby, Hope Greenberg, (notes by) Margaret Tamulonis