Web TV from Dot TV
"The third member of the teaching team is Mark Hayes, who isn't here today because the lucky devil's actually in Tuvalu."
-Professor Michael Bromley to a first year journalism lecture at the University of Queensland, 12.25pm, Monday, February 27, 2006.
"Talofa from Funafuti Atoll in Tuvalu. I'm about three and a half thousand kilometres to the north east of you folks and I'm in the offices of the Real .TV here on Funafuti."
-Dr Mark Hayes, Live via Web Video and Audio Link from Funafuti Atoll, Tuvalu, 12.34pm, Monday, February 27, 2006.
"That was awesome!"
-Overheard from a student leaving the lecture.
The world's first Live Web Broadcast from the Real .TV and Funafuti Atoll, Tuvalu, took place on Monday afternoon, February 27, 2006.
This is the complete story of how it happened, exclusive to Tuvaluislands.com.
(All times Australian Eastern Standard Time, UTC +10)
Early January, 2006
On the back deck of Dr John Harrison's house in inner western suburban Brisbane, John and Dr Mark Hayes were discussing Mark's forthcoming third working trip to Tuvalu in mid-February, 2006, to report on the extreme high tides predicted for the last week of February.
The tides coincided with the first week of teaching in Semester 1, 2006, at the University of Queensland, traditionally when lecturers and tutors meet their students for the first time in classes.
But Mark wasn't going to be at the first year journalism subject's first lecture.
"Hmmm," Mark mused, considering John's idea. "Could be done, but the Real .TV has a tiny bandwidth out of there, and it'd be in the early afternoon their time. I'll talk to my mates there and see what they think."
At the Government Information Technology office, site of the Real .TV (as opposed to all the other .TV presences in cyberspace), Manager Opetaia Simati leaned back in his chair, re-read Mark's e-mail, looked out his window west across the placid Te Namo (the lagoon), and decided that a live web cam broadcast could be done, but bandwidth - the size of the electronic pipe through which the signal could pass - would be a problem.
But he was keen to give it a go.
Another problem was how to show people where Tuvalu is actually located in relationship to Brisbane and the University of Queensland's St Lucia campus, on a bend of the Brisbane River in the inner west of the city.
"Where the hell is Tuvalu?" is a question Mark is often asked when he talks about the country.
Using his iMac and his home Internet connection, with almost as much bandwidth as was available out of Tuvalu, Mark played around with Google Earth, a relatively new feature of the well-known On Line search engine, which uses satellite maps with enormous detail to zoom in on even individual buildings.
Zooming down on the campus, the actual building where the lecture was scheduled on February 27, could be clearly seen and easily identified.
After typing 'Funafuti Atoll' or 'Tuvalu' into the free Google Earth programme, and pressing Return, the image zoomed up, and then out across the Pacific Ocean to settle on Funafuti Atoll.
But the satellite image of Tuvalu was murky and very unclear.
Mark then assembled a PowerPoint display using pictures of Funafuti viewed from space, then from high altitude, both used with permission from Brian Cannon, the maintainer of Tuvaluislands.com, and then one of his own pictures of Funafuti taken from a departing plane in early December, 2004, this last one labled with an arrow pointing to the Real .TV office on the northern western corner of the southern wing of the Tuvalu Government Building: "Dr Hayes is here".
Drawing on his years of experience as a television journalist and producer, Mark deployed this knowledge to mentally plot out several scenarios - Plan A, a Live Web Cam Broadcast; Plan B, recording a video explaining where he was and why which then had to be turned into a computer video file and sent back to Brisbane; and Plan C, sending a series of pictures back to Brisbane after he arrived on Funafuti which could then be assembled into a PowerPoint presentation.
With the Real .TV folks thinking about the Web Broadcast, it was time to call in The Geek at UQ Journalism, Steven R.J. Danieletto, the Journalism School's IT/Technical Support Officer.
"I thought it sounded pretty crazy at first," Steve said. "We hadn't done anything like that before."
"To actually hook up a teacher with his students from three thousand kilometres or more sounded like a pretty original idea.
"It was something definitely we could pull off but we had to think carefully about how to do it," he said.
"The guys at the Real .TV were really nice. We were restricted to text-only chat because of the tiny bandwidth. But it was good working with other professionals from another part of the world," Steve said.
The person who actually had to present the first-year introductory lecture, and call up the Web Cam from Funafuti, was Professor Michael Bromley, professor and deputy-head of the University of Queensland School of Journalism and Communications.
"I thought it was a really good idea," Prof Bromley said.
"It demonstrates how today journalists are totally mobile, sending almost anything from anywhere at any time and aren't dependent on massive amounts of kit and technical back-up and can do it quite simply using wired connections," he said.
"You've always got doubts about whether or not it would work, because I'm a pessimist," he said, laughing. "And experiences in newsrooms say that just at that moment when you want it to work it doesn't work. I always expect there to be gremlins."
"But on the balance of probabilities, I though it would work," he said.
"I thought the students would think, 'Oh, yes! This is how it happens in the real world'. You don't have to get on a bus or into a taxi in Brisbane to go out and do journalism. You can go all over the world and do it," Prof Bromley said.
With The Geek on the case and text chatting with Opet at the Real .TV on Funafuti, and approval obtained from Prof Bromley to develop the idea, the next stage was to resource it.
This was not going to be a simple Web Cam chat, which increasing numbers of Tuvaluans are using, including via very slow speed modems connected to the Government IT Office on Funafuti, but it was also not going to be anything like a live satellite television broadcast now routine on global television.
Rather than low-tech or high-tech, this was going to be an exercise in applied, ingenious 'mid-tech'.
A near-professional quality digital video camera and microphone with a digital signal output was borrowed from the Queensland University of Technology Creative Industries Faculty technical support section because the journalism programme at Queensland University had no suitable cameras.
A laptop computer was borrowed from the UQ journalism school, and custom loaded by Steve Danieletto with all the software he and the Real .TV folks could think might be needed, because, if they forgot something, Funafuti was not the place to source needed gear and software in a hurry.
Dr Hayes, a committed Apple Macintosh user since 1987, had to grit his teeth, 'slum it' as he growled, and do a crash course with Steve on how to use the software on the unfamiliar Windows driven laptop.
Arriving on Funafuti, February 19
Mark arrived on Funafuti on Sunday, February 19, and the next day, met Bennet Simeona and Opetaia Simati in their office in the Government Building.
"It was great to see these folks, good friends of mine, again in person," Mark said.
"I'd met them during previous visits to Funafuti, of course, and we'd been in e-mail, text chat, and occasionally voice communication using Voice Over Internet Protocol (VoIP) using a product called Skype for six weeks before I arrived.
"But nothing beats hooking up with folks like these in person, and working together to make this project happen," Mark said.
An absolutely crucial element in the project was obtained Duty Free at Nausori Airport, north of Suva, just prior to Mark's departure. Without it, the whole exercise would fall apart. Chocolate.
For some totally inexplicable reason, Tuvaluans want all their Palagi visitors to bring them chocolate.
Knowing this vital inside knowledge, Dr Hayes bought up big on chocolate at Nausori, stored it in the freezer at the Filamona Lodge where he was staying, made sure he brought a large package of the stuff with him when he visited the Real .TV that first Monday morning, and, with great formality and ceremony, presented it to Mr Real .TV at their Monday morning meeting.
During the week of February 20, Dr Hayes occasionally text chatted with Steve in Brisbane, exchanged a few e-mails, and discussed aspects of the project's development with Opet and Ben.
A couple of test Web Broadcasts were attempted from .TV to the University of Queensland, which were only seen by Steve Danieletto, using a spare and unique Web address at .TV to prevent hacking or unauthorised access to the data feed.
One worked without a hitch, but another siezed up.
Another issue which could stop the whole exercise in its tracks was if there was a power failure on Funafuti. Nothing anybody could do about this except not even think about it, in case even thinking about it caused the Funafuti Power Station to go into 'sulk mode'.
At .TV, Dr Hayes' borrowed laptop computer was plugged into the network via a high speed ethernet network cable, and the camera was plugged into the laptop through a digital data output socket. Windows Media Encoder was the software taking the data stream from the camera and microphone, compressing it, slowing the picture refresh rate down from real time to about 15 frames per second, and then feeding the stream into the InterNet via the satellite dish on the north western corner of the Government Building.
"Watching the video frame rates while we were tweaking the software and feeds back to Brisbane, I looked like Max Headroom," Dr Hayes, who always notices 'Max Headroom' signs above the entrances to supermarket carparks, said.
The geniuses who invented the InterNet in the late 1960s and early 1970s developed a way for digital data to take several different routes to arrive at the same destination intact. Very simply, this process breaks data such as an e-mail, small text files, or small pictures into 'packets', each packet having its own identifying information. The data stream of 'packets' then obeys the InterNet's rules or protocols, seeking out the quickest way to its destination, which has a unique identifying number or IP address. At the destination, the 'packets' arrive, and are processed by the receiving software into a coherent file which can be read, viewed, or listened to by the slowest element in the process, the 'live ware', or the human being in front of the computer.
Digital video and sound of a quality suitable for viewing involves a massive amount of data, with millions of data packets having to be compressed - the packets contain lots of repetitive data which can be removed at source and replaced at destination in clever ways, to do with psychology, which fool the viewer's brain into perceiving an expanded video and audio file as if it had not been compressed. Audio files in MP3 format work like this, using extremely clever psycho-acoustics.
But Dr Hayes' laptop computer on Funafuti was not especially powerful, the .TV systems into which it was plugged were only a bit more powerful, the pipe through which the data stream had to travel was comparatively tiny, and there were as many as 12 hops the signal had to make from Funafuti, via a satellite to the InterNet service provider in Noumea, then via undersea cable or satellite to Australia, through satellite or fibre optic cable to major Australian Net gateways to the University of Queensland's main Net entry point, which is strongly protected against hackers, and then to the Journalism School's quite powerful server computers.
The traceroute from Dr Hayes' computer in Brisbane, Australia, to one of the Internet addresses for .TV on Funafuti Atoll, Tuvalu:
220.127.116.11 (18.104.22.168), 64 hops max, 40 byte packets
1) 192.168.1.1 (192.168.1.1) 39.260 ms 35.494 ms 39.606 ms
2) dsl-40-1.qld2.net.au (22.214.171.124) 35.470 ms 41.754 ms 34.666 ms
3) fastethernet5-0-20.cor2.hay.connect.com.au (126.96.36.199) 37.477 ms 37.867 ms 34.377 ms
4) ge-0-1-1.bdr5.syd.connect.com.au (188.8.131.52) 33.714 ms 34.204 ms 35.575 ms
5) vlan219.52gdc76f02.optus.net.au (184.108.40.206) 37.574 ms 35.069 ms 45.948 ms
6) gi12-0-0.sb3.optus.net.au (220.127.116.11) 38.548 ms 35.963 ms 36.027 ms
7) newskiesptyltd.sb3.optus.net.au (18.104.22.168) 74.971 ms 65.065 ms 64.809 ms
8) * * *
9) 63-100-216-10.reverse.newskies.net (22.214.171.124) 857.317 ms 604.277 ms 822.494 ms
10) 63-100-216-10.reverse.newskies.net (126.96.36.199) 3614.591 ms !H 3647.370 ms !H 3739.248 ms !H
This shows 10 hops in the connection with a worrying delay at the 8th hop, and, once the signal hits the ISP in Noumea and then onward to Tuvalu, and especially back again, things slow down a lot. Another traceroute might show another path being taken by the data packets from Brisbane, perhaps through New Zealand, with even more hops.
The Journalism School at UQ was where the serious 'grunt work' would be done, provided all the 'packets' of data from Funafuti arrived intact, at pretty much the same time measured in micro-seconds, and could be re-assembled in pretty much the right order. If this didn't happen, Dr Hayes would really look and sound like Max Headroom on a very powerful depressant drug.
Then the fully processed 'clean' Web Streaming Feed of vision and audio had to be fed back into the UQ Wireless Network so Professor Bromley's laptop computer in the lecture theatre could access this feed, and then display the stream, Live, on a large projection screen. Unauthorised users had to be kept off the Feed at UQ because more than one crucial user would further slow the vision and audio replay.
If there's a glitch or problem at any stage along the hops the data stream has to travel, as there are so few paths the data packets could travel across the Pacific, the whole process jams up and falls over.
Knowing all this, but by no means in as much technical detail as The Geek at UQ Journalism and Opet and Ben at .TV, on Sunday afternoon, February 26, Dr Hayes recorded a Plan B video with the help of French documentary producer, Gilliane Le Gallic, on the pier near the Vaikau Lagi Hotel with North Funafuti and Te Namo in the background.
He compressed this on his laptop, turned it into a small Windows Media File, which he sent to Steve Danieletto at UQ Journalism on Monday morning, February 27, using another Net technique used to shift large computer files, called File Transfer Protocol.
Late the previous week, Dr Hayes had e-mailed some Plan C pictures from Funafuti to Steve and Prof Bromley which could be used for a PowerPoint display to which Prof Bromley would speak to explain Dr Hayes' absence from the first year journalism lecture.
At a private function on Funafuti Atoll on Friday night, February 24, Dr Hayes explained to the Prime Minister of Tuvalu, Hon Maatia Toafa, that the Tuvalu InterNet connection would have to undergo 'urgent system maintainence' around 2.30pm on Monday, February 27, and why this had to occur.
In other words, every other user would be dumped off the InterNet Link for about 10 minutes to give Dr Hayes the largest possible bandwidth out of the country for the Web Cam Broadcast. Every electron would be precious, and Dr Hayes needed every one.
Hon Toafa just smiled and laughed, indicating his full support for the exercise.
In Tuvalu, everybody seems to be related to everybody else, one way or another, and it just so happens that Hon Toafa's son in law works for the Real .TV, as did his son before he left for further study in Australia just before all this exercise was underway. Funafuti's a very small place.
During the morning of Monday, February 27, Steve Danieletto and Dr Hayes occasionally text chatted using Skype, but this channel seemed to be having indigestion, with long lags between messages.
This was a Big Worry because Steve was going to text Dr Hayes his "Go!" cue using Skype because Dr Hayes would not see or hear Prof Bromley 3,500 kilometres to the south west in Brisbane, Australia.
Tuvalu Online Home