In a bid to accelerate the growth of the 3D-360 ecosystem, Facebook announced today that it is open-sourcing the specs for Surround 360, the social platform’s high-quality 3D-360 hardware and software video capture system. Now anyone can build and use their very own 360 Surround camera with the Ikea-like manuals that Facebook is providing. The company had first announced about these plans in April.
The company has open-sourced both the hardware efforts as well as the software technology for the 360 Surround camera. Using this, Facebook hopes that developers will develop the required apparatus and content creators will use this to populate News Feeds with 360 video content. The company also says that anyone can now contribute to, build on top of, improve or distribute the camera based on the provided specs.
The equipment that Facebook has listed for the 360 Surround camera is 17 cameras on a UFO-looking stick. The lenses recommended by the company are 4-megapixels each. This will allow content developers to shoot 4K, 6K, or 8K 360 video. The resolution is essentially rated per eye.
According to the social giant, there are a variety of problems in recording and rendering 360 video. These include:
- The amount of data involved is huge. Because we have so many cameras, and because we capture RAW, uncompressed data to preserve maximum image quality, it amounts to roughly 120 GB of data per minute of video (for 30 fps; double that for 60 fps).
- There is little room for error. For mono 360 content, some stitching errors can be tolerated, but for stereo it must be near perfect or it can cause physical discomfort.
- In order to create VR video practically, we need to be able to process all this data as fast as possible, which is often in opposition to the goal of maximizing quality.
Traditional methods and technology do nothing to resolve these issues. So, Facebook came up with its own new ways to solve them. Instead of going with simple old methods, Facebook says that the design of the 360 Surround camera is based on sophisticated geometry.
Essentially the Surround 360 rendering code is designed to be fast so it can practically render video while maintaining image quality, accurate perception of depth and scale, comfortable stereo viewing and increased immersion by incorporating top and bottom cameras to provide full 360-degree x 180-degree coverage and making the tripod pole invisible. The stream won’t be flawless though as some headsets like the Gear VR can’t process or stream that high of resolution fast enough.
The dynamic streaming incorporated by Facebook’s tech puts what you’re looking at in the headset in high definition 8K. The rest of the stream is streamed at a lower resolution. So, if you turn your head while watching a video, you’ll be able to observe the lag in rendering.
If you’re a developer and want to build a 360 camera for yourself, you can get all the info and guides on GitHub. For more in depth info on how the company has designed the Surround 360 camera, head right here.