C ONCEPTUALIZATION OF A H ARDWARE D EVICE TO P ERMIT THE R ECORDING OF L ECTURES IN THE E NVIRONMENT OF P ROFESSIONAL S OCIETY M EETINGS
J
EREMY
H
ERR AND
H
OMER
N
EAL,
U
NIVERSITY OF
M
ICHIGAN E
RIC
M
YERS,
V
ASSAR
C
OLLEGE A
LAN
C
HODOS,
A
MERICAN
P
HYSICAL
S
OCIETY
Proposal:
We propose herein the development of a device to permit the recording of
lectures in the environment of professional society meetings, and a
standardized software architecture for the cataloging and playback of these
lectures, making them available to institutions around the world.
Background
: In our experience recording and publishing high-quality academic lectures on
the web, we have established a process of creating a web lecture, transforming
a live talk into an archived data object that may be viewed by anyone with a
web connection. This is done with both hardware and software tools, and the
process breaks down into three main components: recording the live event
, digitizing and organizing the data, and
publishing the web lecture.
The first component involves recording the video, audio, slides and timing of a
lecture and requires an assembly of appliances, which are individually
available at any electronics store, but which have not yet been integrated in a
useful way. The process of recording a talk currently requires excessive
manpower, rendering it infeasible to record many concurrent talks (as at the
multiple parallel sessions of a national meeting).
The second component entails digitizing and organizing the captured data, and
we feel that the main issue here is the needed adoption of a worldwide standard
for the archival format of this data, which we call a “Lecture
Object”. Although we have set up a working system that creates primitive
Lecture Objects, we need to spend more time crafting a general standard
compatible with existing archival formats and registering our standard with an
internationally recognized bodies.
The third component, publishing the web lecture, has largely been accomplished,
through various software packages that we have developed.
Hardware Description
: To facilitate the first component of the process described above, we
conceive of a hardware device roughly the size and shape of an overhead
transparency projector. It would be therefore compact and transportable, and
big enough to hold the electronic components we desire. The device would be
outfitted with the necessary jacks to capture signals from the sound system,
camcorder and speaker’s laptop. It would have a surface for holding
transparencies, and a document camera mounted above (positioned where the
arm-mounted mirror on a traditional projector would be). The document camera
would photograph the slides and send the image to a video projector, which a
conference room should already have.
Inside the device would be a CPU outfitted with audio and video capture cards,
and connected to a keyboard and screen, so that lecture metadata could be
entered and software maintained. The video capture card would need to be able
to convert a live video feed to MPEG-2 or MPEG-4 in real time, or else a
camcorder would be needed that can output MPEG-2 or MPEG-4 in real time. Both
options exist, so the configuration could be optimized for cost or convenience,
since real-time MPEG-2 conversion is currently only available in high-end
video cards. The innards of the device would otherwise be quite simple,
needing only to collect all the data coming in and supply a comprehensive
timing file containing all events, accompanied by media files. Having such a
timing file eliminates most of the manpower currently required, since matching
up video and slides after the fact is often a delicate and time-consuming task.
Software Archiving Protocol Development:
At the current time, most web users view video over the web using players
from Real Media, Microsoft or Apple. Many other media formats and viewers are
available and their owners are ever vying for greater market share, so that it
is not clear what people will use in the next few years (recent technology news
reports demonstrate the unsure future of formats such as RealMedia and Windows
Media). Therefore we believe that it is essential to archive web lectures
using a format that is non-proprietary and that uses a world-recognized
standard. If the video and slides are also stored in high-resolution (higher
than is currently feasible for web-cast), then the web lectures can continue to
be available far into the future.
Rudimentary work has been done establishing the design of the Lecture Object, a
data object containing video, slides and timing information in a
non-proprietary, high-resolution format. We seek funding to do additional
research and put effort into perfecting and standardizing this architecture,
registering it for example with the W3C. Standards must be established at the
time of birth of a web lecture archive, to ensure its survival and ongoing
usefulness.
Summary:
We envision an opportunity to develop a set of hardware and software packages
that will, for the first time, make it practical for professional societies to
efficiently and rapidly record and upload conference talks to the web,
providing access to thousands of scientists and researchers who are presently
unable to attend talks of interest to them. To accomplish this we require
funding at the level of approximately $100,000 per year for a period of two
years. This amount would cover the cost of one professional, xxxx in supplies,
and xxxx in prototype equipment. We anticipate that further development in the
private sector will be needed to effect the miniaturization required, but that
such funds would be provided privately once the proof-of-principle has been
demonstrated.