Brad Blumenthal, Assistant Professor, Department of Electrical Engineering and Computer Science, University of Illinois at Chicago.
Stephen Borysewicz, Exhibit Developer, Field Museum of Natural History, Chicago, IL
Paul Neumann, Reaearch Assistant, Department of Biomedical Visualization, University of Illinois at Chicago
A new strategy toward these goals in the delivery of museum information is being developed by a group of researchers from the University of Illinois at Chicago, and curators and exhibitors from the Field Museum. This initial Mosaic project includes an online version of the current exhibit "DNA to Dinosaurs" and a subset of the Dept. of Anthropology's collection of Javanese ceremonial masks. The goals of this demo parallel the functions of the Field Museum, both to present information to the general public via exhibitions and to provide educators and researchers access to extensive collections not on display. Use of graphics, sound, animations and interactive games supplies the excitement and educational context associated with museum exhibitions. An important feature of museum displays and collections, as opposed to book, television and traditional multimedia display,is that many artifacts are best understood when viewed in three dimensions. A primary goal of this project is the presentation of 3 dimensional artifacts using 3d display tools, allowing the viewer to investigate the object from multiple viewpoints and to extract dimensional information.
In addition to visible navigational aids, such as the time-line, the on-line exhibit is organized as a linear collection of main pages with a ciruclar detour associated with each main page. The main pages paint the broad structure of the story of dinosaurs; the detours give interesting sidelights on the environment, the people who found and prepared the fossils, and the evolutionary mechanisms examplified by the age of dinosaurs. A main tour button takes the user through the main pages, and can also return the user from the middle of a detour to the most recently visited main page. A "next page" button allows users to explore the detours in an orderly way. Finally, users can wander through the exhibit freely exploring the material in the manner more typically associated with such presentations.
There are several reasons Mosaic documents can be frustrating for the viewer. Many of the creators of the documents have little experience in visual design. But more importantly, the design constraints imposed by the automatic formatting features of Mosaic combined with the time delays caused by the high-bandwith requirements of animations, sound and graphics create formidable design problems. The design decisions of our document tried to balance our goals with the limitations of the media.
Many Mosaic documents end up as long lists of left-justified text and graphics, leaving most of the page blank. Horizontal format graphics of a 600 pixel width were used to better fill the page and accommodate all Mosaic clients. Page length was kept to approximately 1000 pixels to minimize scrolling and downloading time. Navigation and control panels are placed at the top of each page for consistency and visibility. A clever but painstaking technique explained in David Bouma's Home Page allowed us to place images and text in multiple columns in the Triassic Tribune. Adjusting your page width to less than 600 pixels will make the technique apparent.
Picture size was chosen by balancing image quality with download time. In the DNA to Dinosaurs exhibit, images are probably still too large for quick modem access, but further reducing the image size would compromise information carried by the images. In the collection section, smaller images are used for browsing, with larger versions available for closer study.
Reading large quantities of text on CRT's, (as should be apparent if you've gotten this far in this paper), is uncomfortable. In interactive Mosaic documents as in a museum visit, people WILL NOT read large blocks of text. In the online DNA to Dinosaurs exhibit, text was kept to a minimum by using short explanatory captions. In the collections section it was assumed that viewers would be able to print out the text for later reading.
The physical museum exhibit included many video clips: walking dinosaurs, moving continents, and talking heads. These were incorporated into the Mosaic document. Video news reports where the audio was dominant are displayed as still frames of video with audio playback.
Although this is a tiny subset of the actual exhibit and collections, hundreds of hours were spent on scanning, file conversion, writing, content approval, and graphic design. It is clear to the authors that translating high quality museum exhibits into Mosaic experiences can be successful and is worthwhile. The time and resources required to do this, however, implies costs equal to a substantial fraction of the costs of the original exhibit.
Today, U.S. museums and zoos are incredibly popular attractions, with a combined annual attendance larger than that of professional sporting events. But what are people doing at museums? Museum-going is primarily a SOCIAL experience, with most people visiting in family or school groups, or in couples. At the museum, people can eat, visit, touch, listen, and even learn together. And while more and more exhibits are carefully programmed by teams of developers, designers and curators, many people actually use these highly structured environments like shopping malls of information, bouncing from attraction to attraction, raptly absorbed by an interactive or an object one moment, "grazing" absent- mindedly the next.
There should be something familiar about this usage pattern. While the affective experience is certainly different, exhibit visitors' behaviors are mimicked in part by users of interactive computer programs. People tend to skip through branching programs, looking for striking images or sounds, occasionally stopping to read a fascinating tidbit of text or interact with a game or questionnaire. Perhaps the biggest difference, aside from the obvious lack of a third dimension, is in the relative paucity of social interaction among users of many computer programs (this is certainly not true of networked programs, however).
So, is this behavioral resemblance substantial enough to support the translation of a museum exhibit, with its emphasis on affective, interactive and social experience, to a computer program, with its stronger emphasis on cognitive effects? Yes, for three reasons: First, not many people use computers solely as tools for enhancing cognitive experience. Computers are pleasure machines, machines for exploration, and over world-wide nets, machines for instantaneous interaction with millions of potential conversation partners. Second, much of the exhibit contents- sound, animation and video, as well as text and other graphics- can be digitized readily and translated for use on a network program. Third, computer networks serve the same public as museums- the millions of people who view learning as a worthwhile activity as long as it's enjoyable.
The cost of creating visitor-oriented exhibits is often extremely high. The Mosaic version of "DNA to Dinosaurs" represents just a tiny portion of a 21,000 square foot, 6 million dollar exhibit that took four years and dozens of people, from developers to 3-D to graphic designers, to complete. The exhibit serves approximately 1.3 million visitors per year; just over half are from out of town.
On a worldwide network the same exhibit could entertain and educate tens of millions of people. A comprehensive on-line version of the full exhibit would require hundreds more images, dozens of 3-D scans, the selection of video bites from a total of nine two-minute programs already in the exhibit, and the translation of at least six more computer games. We developed our little Mosaic exhibit sampler over several months of coordinated effort, with the help of donated and student labor, and a lot of ad-hoc adjustment of content. A complete program would require substantially more money, and a schedule coordinated with the actual production of the exhibit.
In our project, we wanted the user to browse through a small set of the museum collection interactively. Our first object was a Tyrannosaurus Rex tooth. The tooth's dimensions were measured and entered into a modeling software package (Alias Studio). The resulting curves were skinned together to form a smooth (NURBS) surface which was then written to our internal file format. In the future we intend to import objects by the use of range scanning devices such a Cyberware 3D laser scanner. These types of devices can collect hundreds of thousands of measurements and color values on a real world object within a few seconds. The resulting computer model can be reduced in size by removing near planar sample points.
Our main motivation behind displaying artifacts as 3D computer models was to allow users to interact with museum artifacts as they would with these artifacts in a museum exhibit. Users may view objects from any angle and move closer to the object for a better view. Since the data format for the viewer includes actual 3-dimensional information instead of a 2-dimensional projection, the viewer could, in principle, take measurements of areas of interest. Also, one could vary lighting and shading to emphasize surface information. We hope to incorporate these features in future versions of the viewer.
Mary Rasmussen, Assistant Professor, Department of Biomedical Visualization, University of Illinois at Chicago, has been a faculty member at UIC since 1989, responsible for initiating and teaching the department's three core computer graphics courses. She has worked in interface design, most notably for a program to predict facial growth in children, awarded the first Smithsonian / Computerworld Award for Innovative Use of Information in 1989. Research interests are in facial reconstruction, interactive interface design, and networked distributed hypermedia. e-mail mary@galen.bvis.uic.edu
Brad Blumenthal, Assistant Professor, Department of Electrical Engineering and Computer Science, University of Illinois at Chicago, received his PhD in computer science from the University of Texas in 1990 and went on to a post-doc in cognitive psychology at the Applied Psychology Unit in Cambridge, England. He is currently at the University of Illinois at Chicago where he is applying multi-disciplinary approaches to multi-media design. e-mail brad@ eecs.uic.edu
Stephen Borysewicz, Exhibit Developer, Field Museum of Natural History, Chicago, IL. Stephen works as an exhibit developer at Chicago's Field Museum and at the Shedd Aquarium, creating interactive mutimedia exhibits oriented toward informal learning. He spent the preceding eight years as an editor of medical and scientific publications.
Paul Neumann, Research Assistant, Department of Biomedical Visualization, University of Illinois at Chicago, received his MS at the Electronic Visualization Lab at UIC in 1992. He is currently pursuing his PhD in medical applications in virtual reality. e-mail pneumann@eecs.uic.edu