How Dartmouth Produces Video Podcasts

With an $8,000 investment, Dartmouth's Department of Physics and Astronomy has set up the capability to provide video podcasts for courses that enable students to watch lectures they may have missed or that warrant review. Now, said Lab Manager John Largent, the New Hampshire school is exploring how it can make lecture capture available campus-wide.

Up until the summer of 2004, only a single professor in the department was in the practice of making one of his courses available for viewing after class. The media format was videotape, which had to be checked out from the library.

Video Capture Technology
With seed money in hand and based on the recommendation of other AV groups on campus, Largent invested in a Canon GL2 camcorder. But its MiniDV tape limited him to a single hour of recording time, not enough capacity for the lectures he needed to capture. He switched to 80-minute tapes, and then added the Focus Enhancements FireStore FS-4 portable DTE recorder to the mix. The FS-4 can capture three hours--about 40 gigabytes--worth of video. The tape simply acts as a backup to the FS-4 and vice versa. If the tape in the camcorder runs out during filming, the FS-4 continues recording.

Now Largent can move from one lecture to another without having to download the video store first. Even better, the time taken up by post-production work has shrunk considerably.

He hooks up the FS-4 to his iMac and does a file transfer, which takes about five minutes. The FS-4 saves the lecture in 10-minute clips, so Largent uses Apple's Final Cut Express HD to put the clips together and to edit out "problem sessions," where the professor has spent five or 10 minutes in class having the students work on a problem. That takes another five or 10 minutes.

The big time-consumer, said Largent, is compression. "We put the .mov file into an iPod M4V format." That compresses the file from 11 [gigabytes] to 12 gigabytes down to a file that's 700 [megabytes] to 800 megabytes, small enough to play on an iPod. An hour of lecture took about four hours of compression time until Largent's G5's non-Intel CPU was traded out for an Intel chip; that boosted the computer's processing power, enabling it to do the compression time in two hours.

He said he's hoping to reduce that even more by ramping up the use of Xgrid, Apple's distributed computing utility. "If other computers in the department have that turned on, the computer where all the compression is being done actually uses those computers to do the [work]," Largent said. "It can drop an hour of video the compression down to 20 minutes.... Then we could have lectures up on the Web an hour after the lecture runs."

Likewise, he said he wants to upgrade to Apple's Final Cut Pro, which allows for batch processing. That would enable him to do the edits and compression on multiple lectures simultaneously.

The Eye (and Ear) Behind the Camera
What won't change is the need for his presence--or somebody's--behind the camera. Instructors tend to move around in the room, he said. So somebody has to be there to redirect the camera. Likewise, when the professor writes on the board, demos something or shows slides, the videographer zooms in so that text or screen is readable. "We would try to get PowerPoint slides and inlay those over the video so that when the person was talking, you could see the slide on the screen," said Largent. "But it turned out that that took a lot of time, and the quality of the video wasn't as good. So now, we just tape the PowerPoint up on the screen."

The presence of the person behind the camera is also essential for the purpose of audio capture. Although the camcorder includes a built-in microphone, it tends to pick up noise from the air handlers in the back of the room. So the instructors use wireless microphones. "When the batteries in the mike die in the middle of the lecture, if you're taking it off the line in the back, you lose all your audio," said Largent. "If somebody is sitting there and they realize what's happening, they can pull out the line audio [from the camcorder], and the mike [built into the] front of the camera will pick things up."

Once the recorded lecture is ready for distribution, Largent posts it to Blackboard, Dartmouth's course management system. Students can log in and download files to watch from their computers (the preferred practice), or they can subscribe so the lectures are automatically pushed into their iTunes libraries.

"We've had a great response from students," said Largent. "Fully a third of the hits on Blackboard for a particular course site are for the videos." Interestingly, he said, "We have not seen a decline in class attendance since doing this.... This is how [students] keep up with the class."

Not every lecture in the department is being recorded. Largent said the focus is on the lower level classes--introductory physics and astronomy classes. Nor is every department in the school providing the service. But that could be change. Largent said he recently attended a meeting with participants from the schools of engineering and biology, as well as IT staff, to compare practices and consider how campus-wide lecture capture could be rolled out.

In the meantime, Largent will continue his work, introducing improvements and efficiencies to the system with every passing hour of lecture he records.

To view a clip from a Dartmouth lecture, click here.

Featured

  • person signing a bill at a desk with a faint glow around the document. A tablet and laptop are subtly visible in the background, with soft colors and minimal digital elements

    California Governor Signs AI Content Safeguards into Law

    California Governor Gavin Newsom has officially signed off on a series of landmark artificial intelligence bills, signaling the state’s latest efforts to regulate the burgeoning technology, particularly in response to the misuse of sexually explicit deepfakes. The legislation is aimed at mitigating the risks posed by AI-generated content, as concerns grow over the technology's potential to manipulate images, videos, and voices in ways that could cause significant harm.

  • glowing AI brain composed of geometric lines and nodes, encased within a protective shield of circuit patterns

    NIST's U.S. AI Safety Institute Announces Research Collaboration with Anthropic and OpenAI

    The U.S. AI Safety Institute, part of the National Institute of Standards and Technology (NIST), has formalized agreements with AI companies Anthropic and OpenAI to collaborate on AI safety research, testing, and evaluation.

  • a glowing gaming controller, a digital tree structure, and an open book

    Report: Use of Game Engines Expands Beyond Gaming

    Game development technology is increasingly being utilized beyond its traditional gaming roots, according to the recently released annual "State of Game Development" report from development and DevOps solutions provider Perforce Software.

  • translucent lock composed of interconnected nodes and circuits at the center

    Cloud Security Alliance: Best Practices for Securing AI Systems

    The Cloud Security Alliance (CSA), a not-for-profit organization whose mission statement is defining and raising awareness of best practices to help ensure a secure cloud computing environment, has released a new report offering guidance on securing systems that leverage large language models (LLMs) to address business challenges.