Re: Depth reconstruction in freeviewed images

[P3D <PTWW@xxxxxxx>]

Re: George T. and John B. speculate on my observation of a squash effect
freeviewing a Realist slide of the Grand Canyon as compared to viewing
with the $3 viewer.  George asks what is going on, notes normal freeview
FL of about 250mm; John says he wonders also, speculates I was too far
away.  Jim Crowell reminds us that focusing on the image [surface] while
freeviewing tends to make freeviewed images look smaller.

I did a lot of wondering myself after all this discussion.  Being
nearsighted and freeviewing without glasses, my close focus distance
is around 5 inches, and my normal FV distance appears to be about 8
inches, giving a FL of 200mm, approx, if I am correctly interpreting
how others have described the effective FL when freeviewing (though
perhaps my post earlier today resulted in some new explanation I have
not yet seen?).  It is hard to believe this would qualify as the "too
far away" condition John B. proposed.  So I went back to the lab.
After some experimenting, I thought some more about another comment
John B made:

>Now if Paul was looking at a low-res pair, the disparity might become
>unobservable even sooner since it would be harder for the eyes to pick
>the disparities out of the mush.

Initially I did not give this much thought, knowing my flame shields
would melt in a flash if I were to publicly endorse the notion that a
stereo slide from an authentic Realist camera might be a "low-res pair."
;) ;) ;)

What I *was* able to determine, however, is that the light source that
I used when I made my initial observation was somewhat inadequate, and
may have caused a condition similar to the "low-res" situation John B
suggests.  Viewing close to a bright light, I can resolve significantly
more depth when freeviewing the same stereo pair.  I would rate the
stretch as probably more than I observe with my PT'd 3D, but I would
rate it is as *not* very different from the stretch I observe using a
plain $3 viewer...which comes back to the question about whether simply
expressing the viewing distance in mm produces a measure of effective
FL that is directly comparable to the known FLs of stereo viewer lenses.

Incidentally, with the bright light source, moving father away from the
slide (within my limited range of focus without my glasses) clearly has
a dramatic impact on the stretch, as John B indicated it would.  This
stretch effect appears far more than what woule be predicted by a linear
relationship if 200mm freeview distance is equivalent to 4 times the FL
of the approx 50mm lenses of the $3 viewer.  That is, the stretch effect
is far more dramatic with a 50% to 100% increase of FV distance than it
is in a supposed 300% increase from 50mm viewer FL to 200mm FV distance.
Is this difference due to the focus point distinction Jim C. makes, or
is there something else going on here?

Paul Talbot, freeviewing with more light, but otherwise still in the dark


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