Facebook's annual developer conference F8 meets today and ends. So what is this social platform giant finally revealing about VR-related information? Facebook's Oculus has finally integrated a native tool that allows you to capture and share the Rift experience in 360-degree pictures and video formats. This is the cube mapping technique that Xiaobian once introduced in detail. Developers can now integrate it into Unity and Unreal engines to support Nvidia and AMD graphics cards. The SDK is now hosted on GitHub. But Hocum Lee and Chetan Gupta, software engineers at Oculus, point out that the use of the SDK is not as simple as taking a 360-degree video and putting it together. Lee and Gupta wrote: “To solve the problem, we rethought the way that 360-degree content was created. Generally speaking, the workflow starts with capturing various images, stitching them together, and finally coding them. Previously, we needed to capture content in the game engine while ensuring that we could quickly generate high-quality images on the VR benchmark hardware. Now with the 360 ​​Capture SDK, everything is possible." Extended reading: Depth interpretation of Google's VR video transmission technology: Isometric cube mapping Unlike traditional capture and post-assembly, such a process requires very strict equipment specifications. For this reason, Oculus's 360 Capture SDK uses cube maps. The technology "can run on VR's benchmark recommendation hardware without affecting the experience," which means that you can still meet the demanding 90fps when capturing 360-degree video and pictures. Oculus claims that the device can provide 1080p video resolution for news push viewing and 4K resolution for VR viewing. However, the video is limited to 30 fps, but taking into account the traditional way of combining takes 20 seconds to 40 seconds to complete capturing and flattening 360-degree video, Oculus's cube mapping technology seems to be the best-in-class approach. Oculus said that the SDK plugin is compatible with multiple game engines, including Unity and Unreal, and even native engines. In addition, SDK natively supports graphics hardware such as Nvidia and AMD, which means that most Rift ecosystem devices can capture 360-degree pictures and videos and then share it with VR communities. Lee and Gupta stated that in addition to capturing speed, cube maps have some inherent advantages: There is no geometric distortion on all sides, so each side looks exactly the same as what you see through the perspective camera. This is important because the video codec treats the motion vector as a straight line. This is also why coding is better than bending motion in a rectangular format. Pixels are evenly distributed and each face is equally important. There are no poles in the isometric projection (the poles will contain redundant information). It's easier to project. Each face only maps to the corresponding face of the cube. We realize that we can skip the stitching process and instead use the engine to capture the cube map locally, which saves performance and speeds up. In addition, the quality of the cubic map content is higher than the content of the mosaic, because the former will not lose quality in the process of stitching and converting into spherical projection. Dc 12V Controller,Automatic Switch Charge Controller,New Solar Charging Controller,Rohs Charge Controller GuangZhou HanFong New Energy Technology Co. , Ltd. , https://www.gzinverter.com