Re: questions, questions

[P3D <PK6811S@xxxxxxxxxxxxxx>]

John B writes:

>I'm thinking it's two-aperture 3D that is a little leakey.  8-)  Let me
>rephrase my question: Won't all black objects be at the exact same depth 
>in the Q-dos system?  Sure, they will be occluded by and will occlude 
>colored objects of known depth but it seems like all black objects will 

Try fusing this stereo pair with crossed eyes:       |     /

I noticed this in one of those ascii stereo pics people use for sigs,
and was really surprised at the effect.  Of course both elements are
black!  So it is not the color of the object, but the differing perspectives
of the taking apertures/lenses that gives the stereo effect.

A bowling ball painted flat black and evenly illuminated all round
will look flat to any stereo system including the human one if all
other clues are eliminated.  Fortunately this is a lab situation,
not a real-life one and most objects look different when viewed from
different perspectives - highlights, profile, and relative placement
are some of the ways.

>have zero parallax (left and right images will be the same image in both 

The left and right images will usually NOT be the same.  Put on your
red-blue glasses and look around the room.  Do you see black objects
in stereo?  Which black objects look flat and which have depth?  Why?
Putting on colored glasses doesn't short-circuit depth perception.

>eyes and will have been generated by the entire aperture of the taking 
>lens).

Actually black images don't get recorded at all.  This is a secret
long kept by Kodak which claims to record black better than other film
manufacturers :-)  It is the surrounding colored objects that get recorded
and provide depth clues.

Marvin Jones writes (apologies for re-wording his remarks):

>[... record a stereo pair in black-and-white, view the left image through a
>red filter and the right one through a blue filter...]

And the result will be a strongly-stereo fusion having nondescript color.

These two falsely-colored images can be recorded on one piece of film for
distribution and projection, making this a cheap and easy
solution to the stereo movie question.  Marketing wins over product :-(

>... Viewed through
>anaglyph glasses, the black areas will not be filtered by EITHER lens--
>both eyes
>will see both images without any filtration taking place. The color will effect
>lighter portions of the image, which is why, with a great deal of imagination,
>you can see 3D in a picture shot this way, minimal as it is. This is why
>photographing through colored lenses can not produce a viable anaglyph image.

Black parts are usually just _underexposed_ bits, like the inside of
a closet seen through a partially opened door - not being the main object
of the pic it may come out black.  But its profile (doorway) will look
different from the left and right perspectives.  It might be
hopeless to shoot an entire movie this way, but singles work fine :-) 
Email me your snail-mail address and I'll send you one.

William Carter writes:

>Q-DOS does several things which corrupt it's SL3-D advantages. One of 
>the things it does is space the the red/cyan filters. In doing so it 
>depth encodes only the edges of an object point. This may cause the 
>object point, or an amassed group of object points, to loose the main 
>element our stereo sensitive brain requires to stereoptically perceive 
>depth... self similarity. 

I confess that my homemade versions also have spaced filters (to use
smaller stop for more depth-of-field).  I also confess not to understand
Mr. Bill's thinking here.  And I am doing penance by meditating over
my navel with red-blue glasses on.  ...close the left eye...now the
right one...  Repeat the words: 'self similarity...self similarity...'

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

The red and blue filters are used to overlay a stereo pair on one piece
of film.  That is all.  Anaglyphic movies were mastered in black-and-white,
converting all colors to shades of gray, then overlaid on film using
red/blue filters.  Single-lens anaglyphs preserve much of the original,
'real' color in recording the two images.  The depth is there.
Given its limitations it is a heck of a fun hobby.  :-)

Paul Kline
pk6811s@xxxxxxxxxxxxxx


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