Re: [photo-3d] Apparent vs. actual focus

[Gabriel Jacob <3-d@xxxxxxxxxxxx>]

John Rupkalvis writes:
>Our perception of depth, and depth relationships is very strongly
>  influenced by the characteristics of the subject matter and our
>  logical expectations. This is especially true in the absence of
>  any "real world" depth references.

I couldn't agree more! John goes on to give "real world" depth
experiences he had in a movie theatre with his friend accompanying
him and asks,

>Something to think about.   I am curious if anyone else in the group
>  has ever had a similar experience.

For sure! I think most of us have. I think we resolve relative depth
pretty easily, but have a hard time resolving absolute depth. I think
two reasons for this is due to non-stereo depth cues or lack off, and
conflicting depth cues.

It's interesting your friend and you had different experiences. I share
your friends perceptions. I've noticed some people can see things
popping out of screens very easily and other's need that little
something extra.  I do.

In numerous 3-D movies that I've seen, the comment I've heard was,
how things seemed to be flying off the screen. I usually could not see
it flying "off" the screen, until I saw my first Imax 3-D movie. I speculate
reasons for this are, Imax screens are very large, theatres are very
dark, and the seating is such that the audience in front doesn't
obstruct the screen.

With the screen being very large, there is less chance of seeing
objects touching the stereo window (assuming the stereo window
is set on the screen). Otherwise, if something protrudes beyond
the stereo window, our brain tells us this is not possible and
the stereo depth collapses. Similar with the steeper seating
arrangements, there is no top of the audience head to collapse
the 3-D. The dark theatre also helps to remove any other conflicting
depth cues.

Having said that, if a real world object is closer than the 3-D object
being projected off the screen, this has the effect of accentuating
the 3-D projection. This is an example of relative depth perception
and occlusion.  Same thing applies with objects touching and
protruding the stereo window, see examples below..

Here are some real world examples one can try in the comfort of their
home 3-D lab! I hope everyone has their copy of Olympic SI issue by
now. Take it out and put on the anaglyph glasses. You'll notice the the
tip of the boat is protruding off the page. It's not touching the edge of
the page and it looks pretty convincing. Now stick your finger in it. The
depth will seem to collapse. Now put your finger about an inch and
half (4 cm) off the page near the tip of the boat. Make sure your finger
is a little closer to you than the boat tip is. You'll notice the protruding
boat tip is enhanced.

Another example, here the protruding object is touching the stereo
window. You'll notice on the bottom of the page 38, the splashing
water touching the edge of the page.  It's supposed to be off the page
but is not too convincing. Place the edge of a blank piece of paper
along the the bottom edge of the SI issue. Now raise the paper about
an inch and a half off the issue. The splashing water will now
suddenly seem to be splashing off the page! In this and above
examples the occlusion cues are overpowering the stereoscopic
depth cues.

This wasn't an oversight by SI or David but rather sometimes
choosing the lesser of two evils. In this case ghosting issues would
have made the images intolerable. I've done anaglpyhs where
everything is off the page. In cases like this I hanged the image on
the wall with a frame a few inches above it. This helps dramatically in
viewing the 3-D objects in front of the wall.

For those that don't have the SI issue and since we're talking about
Pulfrich in another thread, here is a Pulfrich example demonstrating
conflicting depth cues. As I mentioned in a previous post regarding
using your mouse to demonstrate the Pulfrich effect, you'll notice the
mouse pointer rising off your screen, if for example you move your
mouse to the right and the dark lens is on your right eye. You'll notice
the pointer hovering over blank areas of your screen or even over text.

Now if you move the mouse pointer to the left, you'll notice the mouse
pointer recedes into the screen only in the blank screen areas and
not over the text. Here again, the occlusion cues are telling you it's
not possible for the pointer to be behind the screen.

2-D or 3-D? It's all relative.

Gabriel