I guess I'll start with Thursday morning's remote camera setup in recounting the latest Adventures at Kennedy Space Center.
Remote cameras are necessary because proximity to the launch can be extremely dangerous. Obviously, on the pad, you'd be a crispy critter. But up to a certain distance (I've heard a quarter mile thrown around), the acoustic energy of the launch alone will kill you instantly. Going beyond that, if for some reason there was an on-pad explosion with a fully fueled external tank, the blast danger area covers a three mile radius around the pad. There's also the issue of the launch plume, which contains hazardous chemicals and drifts with the wind.
Anyway, back to the remote cameras - It's possible there will never be a more diverse group of unique, custom-built, and improvised setups than at a shuttle launch. Everywhere you look, you see something creative, weird and clever.
The camera needs to be able to survive for at least 24 hours in the elements. In Florida, especially this time of year, that means driving rain and strong winds, temperature differentials in the morning and evening, fog, and condensation formation. Even if you're protected from all that, a wild animal might decide to wreck your setup, a wildfire could burn it down, or a spider might decide to nest in the lens hood (That has happened!).
Common housings include birdhouses, hand-built wooden boxes, mailboxes, plastic bins, and bags.
Some groups place dozens of cameras in various locations. At least one photographer I knew spent the entire day out at the pad readying remotes.
The other difficulty is triggering. There are a number of approaches here: Timed triggers, light triggers, and sound/pressure triggers. Some setups combine a timer and a sound trigger - The timer wakes the camera up through the launch window and "listens" for the launch. This aspect likely contributes the most to the failure rate. On average, only about 50% of all remote shuttle launch cameras ever work. That's one of the reasons some groups set up as many cameras as they do.
The setup seen here appears to be using a pressure plate or microphone (Inside the orange bucket) to detect the launch and trigger the cameras:
Here's what my two cameras looked like. Both setups were Canon 30D's using HiViz sound triggers I built at home, all enclosed in Kata bags. It was a very bare-bones setup, with the camera simply staying on and waiting for the shutter circuit to be closed by the sound trigger. If I had more time, I would have set the camera to go to sleep and incorporated a timer to close the "focus" circuit on time and wake the camera back up through the launch window.
The first camera was shooting 50mm and was mounted to a staked-down Tiltall:
That camera failed to trigger completely for reasons I still do not fully understand. It did shoot several frames during a storm, but there were not nearly enough to drain the battery early. The lens was a complete mess as well, with the hood only serving to prevent the front element from drying off. This one was framed tight on the liftoff and I was very disappointed to see it failed.
The second camera was shooting an 18-55 at 18mm on a small tripod placed low to the ground. This one worked and provided the shot I posted earlier.
So there it is, a 50% success rate.
edit: page 2 snipe!