At the time the project began, the simple network delivery of conferencing and teleseminars was fairly well understood. These capabilities were considered valuable for immediate delivery of professional content. But in addition, the capabilities seemed valuable because they also provided a means for capturing and indexing digital presentations by leading experts. These digital presentations were expected to contribute to an expanding archive of previously recorded "best" presentations and seminars, both newly captured and converted from videotape to digital format and made accessible nationwide using the Internet.

Accomplishments

Advanced Network and Services, Inc. (ANSI) and IBM provided the technical support for real time, multimedia delivery and multicast transmission of MIT/CAES content on the Internet. IBM also donated servers capable of delivering video to this effort, while ANSI donated IP priority queuing and multicast technologies that enabled the NSOC/DVF to broadcast CAES seminars via the MBONE and also broadcast the same traffic over the MIT cable network. These facilities were used on a number of occasions to deliver live digital presentations over the Internet.

Experiments were also conducted in digitizing MIT events from both analog and digital sources. As a proof of concept for the potential of combining live MBONE presentations with their simultaneous capture and storage, a set of demonstration MBONE recordings were created and made available on the World Wide Web (WWW). One example of this was an hour long session of Jeff Schiller's Presentation on "Secure System Administration of MITNet" presented and digitized in February of 1995. This complete sample MPEG file is available at the following URL:

http://www.nmis.org/EducationTraining/Schiller/SCHILLER.MPEG (717 MB)

Future Directions

While live broadcasts using the MBONE were relatively straightforward, there were a number of limitations. The restricted use of the system was inconvenient, and the delivery of live presentations could be accomplished in alternative ways. The current rapid expansion of the infrastructure for supporting video could make it more feasible to use live video in the near future. New efforts are being considered based on better technologies for encoding smaller files. More advanced alternatives are also under examination, such as wavelet based encoding/decoding techniques capable of scaleable video distribution. Finally, network routers should soon be more widely enabled to accommodate experimental IP Multicast distribution.