Building a Third-Person Camera Rig - Step 1: Planning

By Will Smith

Will wants to build something silly and awesome for MakerFaire and he wants you to help.

We're trying something new! I'm starting a long-term project and we want to do more than just show you how to make whatever we end up building. We want to involve you in the process, show you how we end up with the plans we have, and (hopefully) get some input from you guys along the way as well. Part two is on the site now.

MakerFaire is about a month away, and I want to build something silly and awesome for this year's show. What pray tell? A real-life third-person camera.

I've wanted to build a third-person camera since head-mounted displays stopped being complete garbage. I have about a month to plan, collect parts, get the work done, and have everything ready for MakerFaire in mid-May.

Photo Credit: Lukas Franciszkiewicz

This has the makings of a fun project on its own, but it's also a fun experiment. I'm very interested to see how different people react to the third-person view. Will it take a long time to learn to walk around or will it immediately feel natural? Will it seem like an out of body experience?

My plan's pretty straightforward. There are four basic components. The camera, the display, the batteries and inverters to power the whole thing, and the frame to hold it all together and make it comfortable to wear for extended periods of time.

First, the frame. I want to mount everything on a frame attached to an internal frame backpack. Everything I've read says not to bother trying to motion isolate the camera, floating the camera on a gimbals or similar mounts creates unnatural camera movement that's much more disconcerting than just putting the camera on a rigid frame. Using semi-rigid PVC to make an A-frame, I should be able to mount the camera at least four feet behind the backpack, so I have a wide enough field of view to see my head and my feet.

Image Credit: Will's Brain

The camera and the display should be a matched pair--after all, there's no reason to pay for an expensive 3D camera if the display doesn't support 3D, and vice versa. While I'd love to get 3D input and output, the 3D camcorders that offer live feed capabilities cost around $1,000--more than I hope to spend.

It seems like GoPro makes the perfect camera with the HD Hero 2. With a 170 degree field of view and the ability to pipe live video out an HDMI port, the HD Hero2 seems perfect for our project. It doesn't have 3D capabilities, but at $300, it's a third the price of 3D cameras and I'm pretty sure one of the models will come with a mounting bracket that will make it simple to hook up to my backpack rig. On top of that, it's rugged and waterproof. This is a good thing.

I'm going to pass on full 3D support, at least for now, but I really want goggles that support an HD resolution. There aren't that many options when it comes to head-mounted displays that render at 720p--we've tested Sony's home theater glasses, there's a Carl Zeiss HMD that won't be available until later this summer, and a bunch of lower-resolution glasses designed for PC gaming. With an $800 street price, the Sony glasses look they're the best option out there, but they're expensive, big, and bulky. They weren't comfortable to wear for long periods of time while sitting on a couch, so I can't imagine how they'll hold up walking around outdoors. The good news is that connectivity between the HMD and the Go-Pro should be straightforward--it's just a simple HDMI connection. The bad news is that power is trickier, since the Sony glasses require AC. Unfortunately, the Sony HMD's control box has an internal transformer--it doesn't use a typical AC-to-DC wall wart, which would be relatively easy to replace with batteries. I'll either have to open the box and bypass the internal transformer, or use an inverter and be prepared to carry extra batteries.

The GoPro has built-in batteries and charges from a normal USB port. I can supplement its battery with a commercially available USB backup battery, or even a solar rig.

I'm not interested in piping live video out of the headset, at least not yet, so that should be all the hardware I need up front. I should be able to simultaneously record using the GoPro while I'm monitoring the live output, so we should be able to show you guys video of me crashing into stuff while I learn to use this rig. I imagine it would look something like the video from this rig, built in 2010:

The next step is to collect hardware, test the field of vision on the camera and the glasses to figure out how far behind me the camera actually needs to be, and head to REI to pick out a backpack to use. We're travelling next week, so I probably won't have an update then, but I'll be back the week after next with a progress report.

In the meantime, if you guys have any solutions for the power problem or suggestions for other head-mounted displays or cameras I should check out, post them in the comments below!

Part two is up now. Read it, if you're interested in seeing how Will's first live tests went.