P3D Re: Dial "M" for Murder a dissapointment

["Greg Wageman" <gjw@xxxxxxxxxx>]

From: Ron Keas <rondonna@xxxxxxxxxxxxx>


>I was devastated when I saw that the film
>was in 3-D, but none of the images came out beyond the screen. All
>images went deep into the screen.
>Have any of you noticed this difference between some 3-D movies and
>others?  I believe it has to do with the original spacing of the two
>lenses when photographing the movie.-  lens close together, like 2
>inches, for objects coming out of the screen, and a wider spacing for
>movie effects like I saw in Dial M.


You must not have been paying attention to the windowing discussions on
this list. :-)

As I understand it, setting the stereo window in a 3D film is usually
done by converging the camera lenses (so-called "toeing in").  The plane
at which the lens axes converge establishes the plane of the window.
Anything behind that plane will be behind the window; anything in front
will be in front of the window.

Many (most?) 3D films were made with the window at "infinity"; that is,
the cameras were converged on the most distant object in the scene.
This was done for a good reason: since infinity points are converged,
the audience's eyes never have to do anything except converge to fuse
anything in the scene, regardless of the screen size.  If it were
otherwise, it is conceivable that, under the wrong set of circumstances,
fusing infinity could require diverging one's eyes, which is difficult
or impossible.

The inevitable consequence of converging on infinity is that most
everything in the scene came out of the window.  I personally find this
annoying for the same reason that window violations in still 3D
photography are considered a flaw: the effect (for me, at least) tends
to collapse at the frame boundary (window edge) because of conflicting
depth cues.  The frame occludes objects so therefore "must" be in front,
yet stereopsis tells you the objects are in front; which cue wins
depends on the individual, but for me occlusion is the more powerful cue
and so the brain tries to make the scene be "behind" the window at the
edges.  The resulting illusion is less that what was intended.

Normally during projection, the frame borders of the two projectors are
superimposed.  This puts the plane of the stereo window at the screen,
and hence puts most 3D films "in your lap" (or face).  However, if the
projectionist separates the infinity points by a small amount (no more
than the average human interocular and hopefully less), the scene, as
well as the window, will recede "into" the screen.

I hope I've got all that right; no doubt some friendly list person will
jump down my throat... er, correct me gently if not. :-)

     -Greg W. (gjw@xxxxxxxxxx)



------------------------------