Comments on: Mathematical Proof 3D movies don’t work

By: Jon

Jon — Sat, 29 May 2010 08:15:39 +0000

It turns out I’m not too lazy to do some calculation. Turns out the distortion is actually more than I thought. For instance, a person sitting just 1 meter in each dimension away from the ideal viewing spot with a screen that is 10m away sees anywhere from 0m (a point which should be on the screen) to 1.414m (or root 2, a point which should be 10m behind the screen) displacement. I guess that is only ~7% error so maybe that is not so bad… The displacement of a point does not depend on the distance between the eyes as long as we assume both observers have the same distance between eyes. The x coordinate of the distorted point does not depend on the y coordinate of the viewer and the y coordinate of the distorted point does not depend on the x coordinate of the viewer. Moving in the y direction has exactly as much effect on the y coordinate of the distorted point as moving in the x direction has on the x coordinate of the distorted point. Those are some interesting things that pop out of the equations.

By: DanS

DanS — Sat, 29 May 2010 04:59:28 +0000

In terms of Jon’s negligible differences I’d say it also depends on the position of the objects relative to the screen, the further behind or in front of the screen the object is supposed to be, the worse the distortion should be

By: Jon

Jon — Sat, 29 May 2010 04:44:48 +0000

I’m too lazy to bother doing the math but my intuition says that the distortion is probably negligible for most parts of the theatre once you take realistic values for distances. That said, the difference between say front row far left and back row far right for instance is probably noticeable. However, my theory is that the movie still makes sense to both observers.

By: Robin

Robin — Sat, 29 May 2010 02:13:47 +0000

Now expand the proof for a curved screen (I believe IMAX uses these). My initial thought is that you will have a region within the theather experiencing similar visual experience. Is it possible to optimise curvature and seating configuration to maximise user experience (most people see the intended 3D representation). It would be nice to analyze a theater and quantify the user experience for a given position within the theatre.