J. Trant

The Getty AHIP Imaging Initiative: A Status Report

A paper presented at the Electronic Imaging and the Visual Arts (EVA) Conference, The National Gallery, London, England, July 1995. Also appearing in Archives and Museums Informatics, Cultural Heritage Information Quarterly, Vol. 9, no. 3, 1995, 262-278.

Contents

1. Introduction
2. The Getty AHIP Imaging Initiative
3. Intellectual Property
   • Museum Educational Site Licensing Project (MESL)
   • The Need for the Project
   • Why Site Licenses?
   • Project Methodology
   • Project Participants
   • Project Goals
4. Education
   • Introduction to Imaging
   • Project Management Case Studies
5. Standards
   • Describing Image Files
6. Conclusion
7. Notes

Introduction

As the success of the EVA conferences has shown, digital imaging has captured the imagination of the cultural heritage community. We have seen dramatic developments in image and multi-media data capture, storage, communication, and delivery. Applications which seemed impossible, or visionary only a year ago are now commonplace. The World Wide Web has brought networked multi-media to the desktop and the home. A burgeoning CD-ROM publishing market has seen a steady growth in art and museum related titles. The fact that several were reviewed in a recent issue of Scientific American shows their move into the mainstream.1

Computer-based kiosks and interpretive tools have become a common part of interpretive programs in museums and galleries. What museum does not have sort of imaging project, in prototype or development? These changes have provided opportunities and posed challenges for the institutions and individuals responsible for the creation of multi-media applications and the development of institutional databases which support them. Possibilities have created new demands. What was once a simple text-based collections documentation system has grown into a multi-media archive, as image, text, sound and moving image are merged into a seamless web of documentation about our cultural heritage. The development and management of these archives has become a significant challenge, requiring an institutional investment in staffing infrastructure and a commitment to a future which sees the virtual museum as a crucial part of the traditional museum's activities.

But where are the guidelines in this new area? The 'problem-set' in relation to digital imaging can seem insurmountable at times. Change seems to be the only constant, and technological advances threaten to overwhelm us. Developing 'standards' in such a landscape may seem a futile exercise, for the time it takes to reach consensus almost guarantees obsolescence. But without some common agenda, we risk the creation of disparate and incompatible systems which are not positioned to take advantage of the full potential of developing network communications technologies.

The Getty AHIP Imaging Initiative

The Getty AHIP Imaging Initiative was launched in the spring of 1994, with a mission to act as a catalyst in the development of the guidelines and standards which would enable the development and network communication of digital image archives. Shared standards and common approaches are essential for the creation of resources to function in the virtual world. Standards enable us to share information effectively and efficiently, without fear of data loss or misinterpretation. Guidelines provide common approaches to problems and help to avoid the syndrome of 'reinventing the wheel'. They enable collaboration and cooperation through the development of consistent approaches to shared problems.

In March of 1994 a group of leaders involved in the construction, management of digital library and museum collections and their distribution over communications networks was asked to define the particular challenges imaging posed for the cultural heritage community as a whole. They were challenged to identify the major barriers to universal and comprehensive access to images and information in art. Consensus clustered around the following areas: Standards, Intellectual Property and Education.2 Standards are essential in order to ensure inter-operability between systems and interchangability of data. An appropriate framework for the administration of intellectual property is a precondition for the ongoing distribution of information and the maturing of the information economy. General education about the potential of digital imaging, including the development of guidelines and best practices, is necessary for the community to move ahead as a whole. Each area raises fundamental questions regarding the creation of a critical mass of digital image and information documenting our cultural heritage.

This framework formed the foundation for the three-year work plan of the Getty AHIP Imaging Initiative, the first year of which has just been completed. Our activities to date have focused primarily in the areas of Intellectual Property and Education. We have also begun to address the more complex issue of Standards. What follows is an overview of our activities in these three areas, which includes a profile of plans for the coming year.

I. Intellectual Property

Few issues have impeded the creative development of interactive multi-media applications with the presence of intellectual property. Just the image of an endless round of letters each asking for the permission to reproduce a particular work, has kept many projects on the drawing board. Alternatively, projects have gone ahead, operating in gray areas of ill-defined copyright law, or working, in the United States, under a broad interpretation of the doctrine of 'fair use'. Neither of these strategies is satisfactory, from either the perspective of the information user or the information provider. Without a common framework of rights, permissions and restrictions, the development of imaging systems is hampered.

Developing a appropriate intellectual property framework will require a complex balancing of the interests of rights holders and the desires of those who use images for teaching, research or enjoyment. Progress in this arena has been slow, particularly internationally, as differences in national law have stood in the way of collaboration. One positive sign, on the national level in the United States, is the production Sample CD-ROM Licensing Agreements for Museums , by the MUSE Multi-Media Study Group.3 Designed to reflect the 'views and opinions of museum professionals" these sample texts with a helpful gloss, offer a starting point for museums faced with negotiating the use of digital images of works from their collection in published, commercial, fixed-media titles. This is a significant first step forward, addressing one of the major ways by which digital images of works in museum collections are now distributed. It reinforces the position that contracts reflect a negotiated position, and asserts the rights that the cultural heritage community may wish to protect.

The networked distribution of digital images requires the development and implementation of new paradigms for intellectual property management. Just as our concepts of geography, and "site" are being challenged by a global network that knows no boundaries - so too are our conceptions of "original" and "copy" dissolving into infinitely replicable reality. Our old models of property translate with great difficulty into network space; we need to look for new ways to manage and license the distribution of information.4

Museum Educational Site Licensing Project (MESL)

The Museum Educational Site Licensing (MESL) Project offers just such an opportunity to explore new means of information distribution. This two year collaboration brings representative museums, colleges and universities together to define the terms and conditions for educational use of museum images and information on campus-wide networks. Launched by the AHIP Imaging Initiative, in association with MUSE Educational Media, MESL will develop methods and guidelines for the educational use of digitized museum materials. A small number of selected educational institutions and museums will collaborate in good faith to define terms of image capture and distribution, and develop guidelines for the use of images and textual documentation from museum collections. Our goal is to develop and test administrative, technical and legal mechanisms that will enable the delivery of large quantities of high-quality images from any museum to any educational institution.5

The Need for the Project

The Museum Educational Site Licensing Project grew out of a sense of shared need and mutual benefit. Both the educational and the museum communities have much to gain from defining a collaboration, which explores the potential of networked communications technology for distributing digital images and information.

Within the educational community demand for access to digital content is increasing rapidly . Campuses are wired internally - often right to the dorm room - and connected externally. Instructional technology departments are making more and more creative use of computer technology for educational purposes, but often lack the rights to use core materials. In some schools, digitization projects have been undertaken without a clear understanding of the rights issues involved. In others, projects are on hold, pending some hoped-for resolution.

As Tom Hickerson, director of the division of rare and manuscript collections at the Cornell University Library and a project site coordinator, said "One of the main reasons we don't have on-line classrooms in the humanities and the arts like we do in the sciences and engineering is simply the lack of information available in digital form." However, much research is still needed to understand how best to integrate digital information into the teaching process. "We aren't going to see the resources devoted to developing large databases of material until we can demonstrate their value. The Museum Educational Site Licensing Project will push us toward those practical uses."

Within the museum community there is a strong awareness of the potential of imaging technology to make collections available. This is paired, however, with a fear of loosing control of the images of the collection if they are released in digital form. Digital images are easily duplicated and altered. They quickly become disassociated from the information they were distributed with. What standards are needed to ensure that images are accurately described? What security measures can be put in place to prevent images from being altered or further distributed. How can the integrity of the digital record be preserved? MESL provides a controlled environment within which to examine these questions.

Museum rights and reproductions departments are daily faced with requests to use images of the collection in digital form, and are often unsure how to respond. What fees should be charged for what uses? What restrictions should be placed on use? And what will fund this new activity in a time of economic restraint? Simply processing requests for the electronic rights to use works represents a major investment in time and effort. New models must be developed to enable the efficient management of intellectual property rights.

Why Site Licenses?

Many models have been proposed to manage intellectual property on the information highway. The majority of these have focused on the development of metering systems, that operate on a 'pay per bit' model. For several reasons, this model does not translate well into the educational or museum community. Monitoring at this level of detail requires an investment in infrastructure and a commitment of systems resources which may be beyond the capability of many institutions.

Charging for the use of information itself seems antithetical to intellectual exploration and learning. Who would be most penalized by limits on access? Both the student struggling to come to terms with the subject, and the scholar striving to develop an in-depth understanding of a particular area of research would generate disproportionately high charges. And who would be responsible for paying them? The individual themselves? The department? The institution? Would each user have a particular allowance? How would such an allocation be made? How can an institution budget for the acquisition electronic resources, when there is neither set charges nor a ceiling on payments in place? How can finite resources be allocated when costs are unpredictable?

The site licensing model which will be tested by the MESL project addresses some of these concerns, and reflects the following principles. Information should be free at its point of use; hidden usage charges should be avoided. (Charging specific access fees would be similar to charging individuals each time the check out a library books from institutional collection.) Costs for assembling collections of electronic resources should be borne as institutional expenditures. Passing fees on to individuals inhibits access to information. Costs need to be predictable - libraries must be able to budget for the acquisition or use of electronic resources - and be reasonable - based on the costs of generating the resource. It should be possible to project costs accurately. Fees should be structured so that the widest possible use of resources is encouraged. Monitoring and security requirements should be reasonable.

The site-license model offers a means of satisfying these concerns. Through an annual subscription fee, an educational institution could gain access to a wide body of quality electronic information about many museum collections. This virtual archive (or selections from it) could be distributed on a campus network and made available to the full campus community. No specific charges would be incurred by the users of the information. Security requirements would be similar to those of maintaining the campus network as a whole, and would not require monitoring at the individual access level.

Site licenses could be administered by a not-for-profit entity which acts as an intermediary between museums and educational institutions. A shared administrative framework would ease both the costs of distribution and the administration of rights. A common agreement would remove the institutional burden of negotiating licensing on a one-off basis - a very labor intensive proposition.

The licensing of material would provide a constant stream of revenue for the museum community, which could be used to add to the available stock of digital images. Educational institutions would have access to a predictable supply of quality images and information about works in museum collections, and would also have the opportunity to act as information providers, making works from special collections and archives available for distribution. A self-sustaining distribution system could be established, which operates for the benefit of both participating museums and educational institutions.

Project Methodology

The core of the Museum Educational Site Licensing Project is a structured test of the distribution and use of museum images and information on campus networks. During the 1995-96 and 1996-97 academic years, digitized images and documentation will be distributed among participating institutions. Technical means of image collection and distribution, and methods for ensuring the security of the digital archive will be tested against requirements defined by the project participants. Representatives will meet regularly to evaluate progress and to develop the needed procedural and licensing framework.6

Project Participants

A competitive process was initiated in the fall of 1994 to select the participants in the Museum Educational Site Licensing Project. Over eighty institutions responded to a Call for Participation; from these proposals, the project management committee7 comprised of experts in the museum, digital imaging, and information networking fields, selected six museums and seven universities to participate in the first year of the project. The Library of Congress was subsequently invited to participate.

The following institutions are participating in the Museum Educational Site Licensing Project:

Museums

• The Fowler Museum of Cultural History at the University of California, Los Angeles
• The International Museum of Photography at George Eastman House, Rochester
• The Harvard University Art Museums, Cambridge
• The Library of Congress, Washington, D.C.
• The Museum of Fine Arts, Houston
• The National Gallery of Art, Washington, D.C.
• The National Museum of American Art, Washington, D.C.

Universities

• American University, Washington, D.C.
• Columbia University, New York
• Cornell University, Ithaca
• University of Illinois at Urbana-Champaign
• The University of Maryland at College Park
• The University of Michigan at Ann Arbor, Dearborn, and Flint
• The University of Virginia, Charlottesville.

Each participating institution has formed an inter-disciplinary project team. Team members are drawn from a broad range professional specializations, from art history to network architecture, digital image capture to intellectual property law. These team members are collaborating in a series of working groups, which form around specific tasks. The current working groups are: Content Selection, Evaluation, Base Measurement, WWW/ Communications, Documentation and Distribution, and Security and Monitoring. In all approximately 150 professionals from across the United States are collaborating on various aspects of the project.

Each participating museum has committed to making a minimum of 500 works of art available in each year of the study. These will be selected by the museums on the basis of criteria suggested by participating faculty, who have been encouraged to propose a wide variety of educational and research uses for the materials. The images and data will be used by the participating educational institutions in at least one course in each academic year, and made available on each campus network. Throughout the project, faculty and student use will be monitored and evaluated by experts in educational technology, providing concrete comparative data about the use of digital images as an educational tool.

Project Goals

At the end of the two year project, model site licensing agreements will be defined to govern the educational use of museum images and information on university and college campuses. Procedures for the collection and dissemination of museum images and information will have been developed and tested and the framework for a broadly-based system for the distribution of museum images and information to the academic community will be in place. In addition, evaluative reports will document the procedures employed in the project. User studies will asses the ways and means by which the images and information were used, and will provide a wealth of information regarding searching strategies, image quality needs, user tolerance levels and the adequacy of image description standards and access vocabularies. Comparative technical data, gathered from each site, will enable an evaluation of the effectiveness of specific architecture's and system topologies. In short, we have an opportunity to assess the educational impact of the availability of a significant body of primary visual information, and design a new way of distributing information from museums to universities.

Creating this new set of relationships will require a re-examination of the role of information provider and information consumer. Key in this reengineering exercise is an economic study of digital museum and library operations. Little is known about the real costs in the ongoing distribution of images and information over networks. Within MESL we have an exceptional opportunity to build our knowledge of the issues involved in delivering large electronic data sets, in different technical and administrative environments. This will enable us to develop and test economic models with a hope to creating a more general framework - a self-sustaining mechanism that will enable the educational use of material from any museum by any educational institution. In order for this model to count as a success, however, we must develop a solution that will work in numerous national jurisdictions. The MESL Project is willing to share its expertise with any organizations who might wish to mirror the model in their own countries, and we look forward to the possibility of international collaboration.8

II. Education

Introduction to Imaging

It is easy to forget, as we embrace advancements imaging technology, that this is an arcane and elitist world, still inhabited by a few, who speak their own language, and often communicate in code. Demystifying the world of digital imaging was our goal in the production and publication of an Introduction to Imaging: Issues in the Construction of an Image Database.9

This booklet introduces the technology and vocabulary of digital imaging as they relate to the development of image databases depicting works found in museum collections. It illustrates the choices that must be made when images are captured, and outlines the areas in which institutional strategies regarding the use of imaging technologies must be developed. It does not offer "the answer" to the question, "which image database software should I use?" for such a reply is as elusive as the Holy Grail. What it does provide is an outline, which will enable those in Management to appreciate the potential of digital imaging, and provide those in the initial planning stages of an imaging project an overview of the issues they will face, and a means to structure their research. Throughout the text, a strong emphasis was placed on identifying the core of a Research Agenda, listing Other Resources where information can be found, and developing a detailed Glossary and Bibliography.

We have also made the text of the Introduction to Imaging available on the World Wide Web (http://www.getty.edu/research/institute/standards/introimages/01-Intro.html ). This transformation of a published document into a resource site required a rethinking of the structure and sequence of the materials presented. Issues arose in a number of areas relating to the nature of the Introduction to Imaging and the content that it communicates; to the design and implementation of an AHIP presence on the WWW; and to questions of the nature and direction of developments in the WWW itself. We also had to struggle with the limited typographic capabilities of HTML. Without any control over the final delivery platform, design becomes an exercise in balancing communications goals within the restrictions of communications networks.

As the hypermedia document took shape, preliminary strategies for addressing issues in structure and delivery of content were adopted. The construction and production of the WWW version of the Introduction to Imaging became a practicum for the subject matter it introduced-it is a network-accessible document, incorporating images, about making digital images accessible over networks. As such, it provided an opportunity to use the medium to communicate aspects of the message-for example, to demonstrate the implications of file size on network transmission, rather than to describe them. This self-reflective nature added an additional dimension to the design and implementation of the site.

We will be maintaining the on-line version of the Introduction to Imaging, and welcome comments and suggestions about how it may be enhanced and improved. Please use the on-line comment tool to send us your thoughts.

Project Management Case Studies

In the next year we plan to build on the framework offered by the Introduction to Imaging, by developing a series of Project Management Case studies. Our goals in this area are tripartite: first, to build knowledge of the implications of particular project management strategies in the development of imaging systems; secondly, to identify critical moments in the project development life cycle of imaging systems, when choices are made which influence the future direction of the project; and thirdly, to assess the effectiveness of certain technologies, within defined circumstances.

This planned study of imaging project management will highlight the "best practices" in the management of imaging systems. It will outline the phases in the development of an imaging system - from requirements analysis, through technology choice, to implementation and assessment. Each project will be profiled in a similar manner, in order to allow for comparisons. Particular attention will be paid to the requirements analysis phase, and to the development of a framework to assess the 'fit' between imaging requirements and technologies.

As we are still in the process of finalizing the projects to be used as case studies, we would welcome suggestions for inclusion. A preliminary report on our progress in this area will form part of a pre-conference project management seminar that the Imaging Initiative is sponsoring at the International Conference on Hypermedia and Interactivity in Museums conference in San Diego this coming October.

III. Standards

Shared standards must be the foundation upon which the virtual museum is constructed.10 As the cultural heritage community develops multi-media archives which document our collections, and takes advantage of telecommunications networks to share information with our various communities, we will rely on common standards to ensure that information maintains its meaning and its value as it moves across networks and among systems. Reusable digital archives documenting museum collections must be constructed around a framework of standards to safeguard their integrity and longevity. No single "standard" will be sufficient. Multi-media applications include diverse data types, and take advantage of a range of system capabilities. Each of these areas is governed by a particular standard or set of standards, either developed within the international standards community, or within industry.

Standards apply as information is captured, stored, displayed and distributed. The challenge for the cultural heritage community is not to govern the development of technical standards, but to understand their appropriate use. We need to build our understanding of the suite of technical standards so that we can make choices that reflect our needs, and so that we can focus our development efforts on areas of particular concern. The Imaging Initiative has identified two specific areas where the cultural heritage community needs to participate in the development of standards: image documentation and image quality. If we wish to ensure that we will be able to use digital image files in the future, and accurately interpret the information they contain, we must construct them according to consistent guidelines, and document their intellectual and technical characteristics.

Describing Image Files

The Imaging Initiative has participated in discussions of image documentation requirements within the context of the Coalition for Networked Information, and the Consortium for the Computer Interchange of Museum Information. In the fall of 1994, a preliminary framework was presented for comment, which raised the following questions: What information should be recorded about digital image files? Where and how should image files be documented? What information must be placed in the image header, and what can be placed in an accompanying text record? How should that text record be fielded? How should the relationships between image files and their accompanying text records be managed?11

A number of general kinds of information about a digital image were identified:

1)

Information required to view the image, including type (bit-mapped, vector, video), format (such as TIFF, GIF, JFIF), compression scheme (such as JPEG, LZW, QuickTime) pixel dimensions and dynamic range, CLUT and color metric (CMYK, RGB);

2)

Information about the quality and accuracy of the image, including the source digitized, the source image type, source image identification and the institution responsible for creation of the digital image (this could be a series of recursive relationships when images are derived or scanned from other images);

3)

Information about the scanning process, including light source (full spectrum, infrared), resolution, dynamic range, type of scanner (for color correction), date of scan, the identification of the scanning personnel, a record of image manipulation, (cropping, color balancing), and the addition of digital signatures or other methods authentication;

4)

A description of how an original is depicted in a surrogate image, including mention of perspective, position, orientation, aspect, and linking between various views of the same original, or different versions of an image (browse, high-resolution, medium resolution) derived from the same scan;

5)

A description of the work depicted, according to a known content standard, such as AACR2 or the AITF Categories for the Description of Works of Art;

6)

Rights and Reproduction Information, documenting the copyright of the original, the surrogate image, and the digital image, and including the name of the rights-holder and possible use restrictions (on viewing, printing, or reproduction);

7)

Information about how to locate an authentic copy of the image, recorded in a form such as a Universal Resource Name/Number or Universal Resource Locator.

This preliminary schema needs further examination and refinement. Each of the areas identified above needs additional definition and clarification. We need to understand which data should be carried along with image data, as part of an image file, and which should be stored separately. This must be done to satisfy two critical needs. As image files move around the network it is essential that they carry enough information with them to identify their contents. And as image archives are constructed, it is essential that we record the circumstances of their creation, so that we can evaluate their contents in the future.

Image quality standards are also essential if we are going to create digital image archives of lasting utility. While we have historically focused on capturing as much digital data as possible, we must develop realistic guidelines for acceptable quality which move beyond resolution, and dynamic range as measures. If we wish to use digital surrogates for research and teaching, we must build our understanding of the functions for which the image is to be used. Only then will we be able to make sound judgments about appropriate amounts of information to be delivered under certain circumstances.

Our assessments of image quality need to accommodate more complex concepts of color fidelity, the quality and photographic intermediary, and the characteristics of the original work depicted. A useful model here is Digital Resolution Requirements for Replacing Text-Based Material: Methods for Benchmarking Image Quality, recently published by the Commission on Preservation & Access.12 We also hope that the work of the Museum Educational Site Licensing Project will enhance our knowledge of the use of digital images in classroom teaching, self-study, and research.

Conclusion

There is no shortage of issues to address in the area of digital imaging. For example, as the paper "Research Topics In Image and Multimedia Retrieval," prepared by Donna M. Romer for a discussion AHIP is hosting on the Research Agenda for Humanities computing, outlines we have many further challenges the integration of images into what is a primarily text-based world of information search and retrieval. Resolving these questions will require our collective efforts; the Imaging Initiative looks forward to future collaboration with those who share a common interest in their resolution.

Communications networks should provide us with the greatest motivation for this collaboration. Not only do they enable professionals in disparate geographic locations to work together, but they provide the platform which makes the virtual museum a plausible reality. It is now possible to consider merging data and images from a number of networked sites into a resource which reflects the interests of a particular student or researcher. As more intelligent desk-top tools move the Web away from its current configuration as a set of 'destinations' into a fully functioning virtual museum, where it is possible to retrieve representations of museum objects in conjunction with rich contextual data, rather than simply viewing them in a predetermined manner, our collective vision, of a distributed knowledge-base of our common cultural heritage comes into focus.

Notes

-- Note: Links to the MESL project site are no longer active, as this site is no longer maintained.

1

Ben Davis. "The Gallery and the Machine". Scientific American . May 1995.

2

A detailed report of the discussions at this meeting, entitled "Initiative on Electronic Imaging & Information Standards, Meeting Report, March 3-4, 1994" is available from the Getty Art History Information Program. Contact Michele D'Amico, AHIP Imaging Initiative.

3

Muse Educational Media. Sample CD-ROM Licensing Agreements for Museums. 1995. Available from Muse Educational Media 1 East 53rd Street, 10th floor, New York, NY, 10022-4201.

4

See for example, Esther Dyson. "Intellectual Value". Wired. July 1995, 137-141, 182-4.

5

Background information on the Museum Educational Site Licensing Project is available by ftp from the Getty Art History Information Program: ftp to ftp.ahip.getty.edu/pub/mesl , log in as 'anonymous' with your email address as the password. A profile of the project can be found in the publication of the Museum Computer Network: . J. Trant. "The Museum Educational Site Licensing Project." Spectra, Spring, 1995.

6

Two full participants meetings have taken place to date. Reports are available on the AHIP ftp site.

7

The MESL Management Committee is comprised of: Maxwell Anderson, Director, Art Gallery of Ontario, and Chair of the Information Technology Committee of the Association of Art Museum Directors; David Bearman, President, Archives and Museums Informatics; Howard Besser, Visiting Professor, University of Michigan; and Clifford Lynch, Office of the President, University of California. The Project Executives are Jennifer Trant, Manager Imaging Initiative, Getty Art History Information Program and Geoffrey Samuels, MUSE Educational Media.

8

Further information about the Museum Educational Site Licensing Project can be found through the Getty Art History Information Program's World Wide Web site.

9

Howard Besser and Jennifer Trant. Introduction to Imaging: Issues in Constructing an Image Database. Getty Art History Information Program, 1995. Also available online at http://www.getty.edu/research/institute/standards/introimages/01-Intro.html

10

The framework presented here draws on that presented by Costis Dallas, Director of the Foundation for the Hellenic World, at the meeting of the International Council of Museums, International Committee on Documentation, Stavanger, Norway, July 4, 1995.

11

J. Trant and H. Besser. Describing Image Files: The Need for a Technical Standard, Coalition for Networked Information, Fall Meeting, Orlando, Florida, November 30, 1994.

12

Anne R. Kenney and Stephen Chapman, Tutorial: Digital Resolution Requirements for Replacing Text-Based Material: Methods for Benchmarking Image Quality, Commission on Preservation and Access, April 1995.

1995/07/15