It is my understanding that we can perceive depth (3 dimensions) with our eyes because each eye provides a slightly different perspective of the world. Knowing this, we can apply it to regular 2d videos. At one point or another some part of a video will be viewed from a range of perspectives, for at least a number of seconds. Like when a person turns their head as they speak. Or the camera rotates around a character, the fly-by of a plane, a car, or spaceship. All this requires is for the camera to move along giving us different perspective across the length of a scene in a video, or for the subject in the video to move in such a way to provide perspective information.
So we are using the video a few frames ahead to feed this secondary image to your eye, while you’re other eye will be fed the video a few frames behind the other image.
What do you think?
In principle it would be possible, but the movement would have to be very uniform and the uses would be quite limited. It only takes a fraction of a second to turn your head, so your brain might not even register the 3D effect. Furthermore, when you speak your lips will move and your expression will change, which would be a problem if using two different frames.
The fly-by of a plane would not work, because the object needs to be in the same spot for both eyes.
Personally I would not like to wear the 3D glasses for an entire movie just to be able to see a few glimpses of 3D.