"Video is an increasingly popular means of sharing our experiences and connecting with the things and people we care about," explained Facebook software engineers Evgeny Kuzyakov and David Pio, in a blog post on Thursday. "Both 360 video and VR create immersive environments that engender a sense of connectedness, detail, and intimacy."
Both Pio and Kuzyakov admitted and then outlined the challenges presented by delivering 360 video at scale while maintaining quality that would satisfy both content makers and end users.
For example, the duo noted that the standard equirectangular layout for 360 videos flattens the sphere around the viewer onto a 2D surface, which ends up warping images.
The remedy for delivery on the News Feed, they continued, was to remap equirectangular layouts to cube maps. As for VR, Facebook went with more of a pyramid-like scheme in an effort to keep file sizes to a minimum.
"Cube maps have been used in computer graphics for a long time, mostly to create skyboxes and reflections, and we wanted to bring the same capability to anyone who wants to stream high-quality video," they explained.
Facebook is open sourcing the code for its custom 360 video filter for cube maps, which is being made available via GitHub today.
Work with optimizing 360 video for VR is ongoing, and the software engineers revealed Facebook is developing a machine-learned cost function to more efficiently call up video snippets and predict head orientation.