Wonder what a 360° video is?
In this page we describe what it is and what is involved in the production of a 360° video.
Cameras, as we know them up till now, are characterized by a front, a back, a field of view and they describe a portion of what happened in front of them giving the spectator a point of view. The director decides what to show and what not to show to his/her viewers. Essentially when we watch a video we hear a story.
A 360 camera is a new type of camera that has no front, no back, no left and no right. It records everything that happens around it and gives the spectator a point of presence. The viewer is immersed into the scene and it is him/her to decide where to look. When we watch a 360 video we live the scene.
The output of a 360 camera are a bunch of videos: one from each camera that composes the 360 array. To transform them into a spherical video you need to go through the stitching phase, one of the hardest part of the workflow.
In this phase you need to pick a frame from each camera and join them into a single merged one. This process must be repeated for every frame that composes the video. The difficulty in doing this is that the human eye is very capable in detecting anomalies and errors and they are felt as unnatural.
Once you have the spherical image you need a way to represent it. How to draw the surface of a sphere on a two dimentional plane is a long dated task that man solved many centuries ago. Mercator, in the 16th century, invented his famous projection that we still use today to draw the surface of the earth on two dimentional maps.
Unwrapping the surface of a sphere on a plane is an impossible task and some distortion must occur: things near the equator maintain their proportions while things near the poles get more and more distorted.
For 360 videos we use a similar approach and we use a projection called equirectangular. In this way we can treat a 360 video almost as if it were a standard 2D video.
Watching a spherical video in equirectangular projection is not very exciting: to watch it in the proper way we need a dedicated player that understands the equirectangular format and provides extra controls. We can use a variety of devices to watch a 360 video:
The player must transform the spherical image into something closer to human sight. Since we have a spherical image that represents what happened around the camera the idea is to place the viewer right where the camera was: in the center of the sphere. The viewer can now look around and see how the sphere looks like, seeing a portion of the scene at a time as if it was where the camera was standing at the moment of shooting.
If the viewer does not move from his/her position the illusion is perfect.
The way in which the viewer controls where to look depends on the device used.