October 13, 2005
HDTV, live from the seafloor
With clever engineering and being in the right spot, under the right satellite, UW oceanographers working with Computing & Communications and the ResearchChannel became the first team in the world to broadcast high-definition video from the seafloor to selected sites around the world Sept. 27.
At the same time, television viewers and computer users tuned to UWTV and the ResearchChannel were receiving, most in standard definition, their first live look at sites on the Juan de Fuca Ridge, the Yellowstone of the deep that’s just off our coast.
High-definition video, which provides seven to 10 times the clarity of standard definition, requires signals that are considerably faster and richer with data. Along with the challenge of achieving such a signal from the UW’s research vessel Thomas G. Thompson, other struggles ranged from a shipboard transmitter that ceased working when the expedition had been at sea five days and the need to call in a favor the last week of the voyage so the TV crew necessary for the live broadcasts could be ferried out to the ship — a trip that took 24 hours in increasingly unfriendly seas.
That’s because a storm was brewing that threatened the opportunity to broadcast live and meant the thousand-plus hours of engineering would have to pay off during some future expedition.
“It was alarming,” says Christine Ruiz, UWTV producer. “So much work went into preparing equipment and routing, into the scripting and the one thing that looked as if it would trip us up was the weather.”
With weather forecasts in hand, co-chief scientists John Delaney and Debbie Kelley and UWTV engineers turned what had been planned as a rehearsal Sept. 27 into the real thing, complete with live video broadcast from the seafloor in high definition. Broadcasts announced for Sept. 28 and 29 were aired as planned, during which the scientists were live from the ship but the video from the seafloor was pre-recorded from earlier in the expedition.
Stormy weather made it both unsafe to launch the remotely operated vehicle that carries the high-definition underwater camera to the seafloor and meant possible signal loss as pitching seas made it difficult for the ship’s antenna to stay locked on its satellite.
With or without bad weather, that connection from ship to land via satellite has been the mechanical reason no one has been able to do this before, according to Mike Wellings, director of engineering at UWTV.
For one thing, even large research vessels like the 269-foot Thompson don’t have room for antennas much bigger than about 8 feet, which limits the size of amplifiers or transmitters.
For another, from many locations at sea, it’s not possible to push enough of the data needed for live high-definition broadcasts through most satellites, the pipeline just isn’t big enough.
“We were fortunate to use a satellite that’s one of the most powerful out there and one already used by the ResearchChannel,” Wellings said. The ResearchChannel broadcasts to a national audience, available to subscribers of the Dish Network, with programming supplied by educational institutions like the UW.
Sitting at the dock in Portage Bay, the satellite dish on the Thompson was able to push 20 megabits of data, which is the rate used for high-definition broadcasts by Seattle stations that offer them, such as KOMO 4.
Once at the research site 200 miles off the coast, however, that rate dropped to 15 megabits, enough to accomplish what was needed. That’s because the powerful Galaxy X satellite being used is focused on North America and its reach drops off out at sea, about where the scientific expedition would be underway.
“Until we actually got out over the Juan de Fuca Ridge, we didn’t know if it would work,” says Dave Robertson, the UWTV engineer most heavily involved in this expedition. Among other things, Robertson helped upgrade the satellite dish on the Thompson, tracked down information needed to get the right high-definition camera and went to sea for 16 days. “This whole thing took over my life last May,” he says.
For this project, UWTV also was striving to send high-definition video by satellite, not in the standard broadcast format used by television, but in the language of the Internet.
“We used an IP-based transmission model because we wanted to prove we could do it and to push the boundaries a little,” Robertson says. The UW already uses IP-based transmissions (the IP stands for Internet Protocol), for campus networking and phones.
“Since UW is well provisioned with high speed networks, both on and off campus, it provides a good test bed for this kind of experiment. Eventually video can be distributed on campus networks with existing infrastructure.”
The same video can be distributed to scientists and educators at other places via the Internet, once the infrastructure is in place at those locations.
“This IP-based feed is an important step in transforming the way research is conducted,” says oceanographer Delaney. “It is the result of an exciting collaboration with resources from the ResearchChannel, National Science Foundation and the W.M. Keck Foundation.”
Last month’s live broadcasts offer a preview of future transmissions from seafloor observatories, such as NEPTUNE where scientists intend to wire 150,000 square miles of seafloor off the west coast of Canada and the United States. Arrays of instruments, cameras and robots, interlinked by fiberoptic cables from land, will be used to study such things as erupting volcanoes, migrating marine mammals, deep-sea ecological change, encroaching pollution and the generation of tsunamis.
The television broadcasts Sept. 27, 28 and 29 were available only in standard-definition — the same quality as regular television. But a Windows Media version of high definition video was available to anyone with a reasonably high bit rate connection across the world.
The high-definition broadcasts were sent to selected research groups and sites in six countries, capable of handling the high bandwidth data, and which included the iGrid conference in San Diego for the world’s high-end iGrid computing and networking groups. Each site had its own specifications.
“No other institution could have done this in house,” Wellngs says.
Again, the video was embedded in an IP stream for transmission. Sending it as uncompressed video was facilitated by Northwest Gigapop and Pacific Wave, two regional data transfer centers that are helping remake the Internet so it can provide pathways a thousand times faster than today’s typical commercial Internet connections.
While the C & C team was preparing for high-profile broadcast demonstrations of live high-definition broadcasts here on land, Robertson on board the Thompson was witnessing the results of his own “high-profile” demos, the ones for the scientists and crew.
“We had three HD screens placed in different locations around the ship. The images coming up from the seafloor were just so compelling and engaging. I remember working in front of one of the screens while tube worms were being shown. One of the biologists leaned over my shoulder all excited saying, ‘Look, you can see their gills.’
“HD isn’t just about gee whiz, it’s a research tool.”
In the booth: Stormy weather threatens broadcast
Tension ran high in the UWTV control booth on campus as stormy weather 200 miles away repeatedly caused the satellite signal to cut out during two of the three live broadcasts with shipboard scientists last month.
The losses turned out to be momentary — resulting in briefly pixelated images and lost bits of audio, says Christine Ruiz, UWTV producer and director choreographing the live broadcast programs. But the threat of losing the signal altogether because of the storm was very real.
Ruiz was the picture of calm last week at the edit suite console, helping clean up the glitches and preparing the three programs for rebroadcast in coming weeks on the ResearchChannel and UWTV. Both stations are available on campus cable channels 12 and 27 respectively. Check for dates and times at http://www.researchchannel.org/ and www.uwtv.org .
The Sept. 27 live broadcast is also available as streaming video on the Web at http://www.researchchannel.org/visions05/multimedia.asp#video.
Aside from what Mother Nature threw at the broadcasters at the last minute, Ruiz says another major production challenge was how to handle the communications that go on behind the scenes of any broadcast; for example, how the producer talks to the speaker while also giving directions and cues to the videographer, technical director and audio operator. Traditional shipboard communications involve low-bandwidth satellite transmissions that cause noticeable delays for anyone trying to converse by satellite phone and e-mail is batched and delivered enmasse only a handful of times a day.
So, the month of satellite time purchased to transmit video also was used to provide speedy telephone and e-mail communications via an IP, a reliable Internet-based solution. Computing & Communications phone and e-mail experts worked out the complex routing required; there were, for instance, three IP phones used on the Thompson, and the result was seamless during the broadcasts, Ruiz says.
On board the ship, UWTV engineer Dave Robertson said one of the biggest production challenges was faced by the scientists. They were trying to do an extremely complicated agenda of science, and prepare for live broadcasts.
“The chief scientists John Delaney and Deborah Kelley did a great balancing act,” he says.
- Sandra Hines