Chapter 29: Stereoscopic recording
Excerpt from Chapter 29: Stereoscopic recording, page 150:
...If you really want your 360˚ video to be deeply immersive, transporting the viewer to another location that feels real and authentic, you must shoot not just “in the round,” but stereoscopically (aka in 3D) too.
Shooting stereoscopic 360° video requires different, specialized cameras that capture two images from slightly different angles (just as human eyes do) for each frame of video. Later, these two images are combined in an HMD to create the illusion of depth.
For the purposes of 360˚ video, there are two kinds of stereo rigs:
A zone-based stereo camera mounts pairs of camera sensors positioned with a precise (and sometimes adjustable) interocular distance.
An omnidirectional stereo camera that uses an array of cameras evenly spaced and generate the stereo separation programmatically. (Image courtesy YI Technologies)
- Zone-based stereo camera: whereby a series of paired cameras are mounted together. Each pair of cameras captures two side-by-side images (left and right) and then the left cameras are stitched together, the right cameras are stitched together, and the two stitched files are combined to create stereoscopic output.
- Omnidirectional stereo camera: This type uses a single camera body arranged in a ring with multiple lenses (and multiple image sensors) that provide enough overlap between adjacent lenses to allow for a computationally generated stereo image in all directions.
Unfortunately, recording in stereo can make shooting a lot harder. For starters, stereo requires a more complex, usually larger camera rig. Managing the media from a mono 360˚ camera is complicated enough. Stereo doubles the number of camera sensors, making media management exponentially more complex (as described in Chapter 32: Ingesting footage). ..