Normal Base Macros

[T3D <LeRoyDDD@xxxxxxx>]

John Bercovitz said:

"So what's your macro philosophy?..."

-----
     Everything will work out well in the long run. Micro philosophy is work
hard and smart to make the macro philosophy be true. :=)

     But seriously and in context, do what you need to achieve the effect you
are seeking. This takes random and/or methodical experimentation to see what
effects you can achieve and even to decide which to seek.
     The geometrical aspects of the theory of 3D are a place to start, but
the brain can and will often overide what the theory indicates "must" happen.
-----

"I was thinking of taking some slide bar macro/tabletops but first I have to
ruminate, as usual (that's the fun part, aside from looking at the results)."

-----
     For TABLETOPS, one approach is to scale all dimensions the same,
including interlens spacing.

     Unfortunately, your camera doesn't scale. This isn't too bad for 1:12
tabletops. Your SLR is roughly equivalent to a 12x18 "viewcamera" without
movements(for the regular lens) that only stops down to f22. Your scaled lens
spacing is about 5mm.
     For .03mm circle of confusion, a 35mm lens focused at 1.5 feet yields
about 8.5 in. depth of field starting at about 14.5 in. So you could do a
"realistic" 3D tabletop of a dollhouse room or a restricted depth outdoor
scene. Most people aren't going to construct a tabletop scene including scale
stereo infinity(25 to 50 ft. actual), anyway.
     Depth control techniques will work at this level of miniaturization,
though, to fight limited depth of field and still afford realistic depth.

     HO scale(1:87, I think), though, is much tougher. Interlens separation
of about .75mm is just the start of the problems. Think scale hyper in this
situation. A normal lens separation scales to about 18 ft. Just fine for an
"aerial" hyper from 500 to 1000 scale feet(5 to 10 actual).
     To get closer to "realistic" depth, I think you need to go to pinhole
techniques.
-----
     
"I could  use the Spicer-Bercovitz maofd formula and take viewable macros
from fairly close up using a 65 mm stereobase. Although this would be
realistic, I don't think it would look so hot because when you mount them,
you'd probably mount to the window and a standard window is a ways away.
 It's far enough away that I think the vergence clue would intrude and tell
you it ain't so."

-----
     Not a problem for me. I can see my fist at six inches from my nose, I
can photograph it with a regular(almost) 3D camera. My experience is that in
the viewer and in a dark projection situation, if "ortho" conditions aren't
violated too much, *the brain experiences the subject/scene where it ought to
be based on the recorded relative deviation of points in the scene.*
     I have decoupled the lenses on an RBT X2B, Tokina 28-70 macros, and shot
close up at the closest macro that can be set.
     Decoupling the focus and setting it at 2.3 ft. gives an image about 3 to
4 in. high and mountable in half frame to Realist width mounts. Taking is
from about 12 inches or so and the actual lens separation is probably about
40mm due to taking through the "inside" halves of the lenses.
     I sense the scene at about 6 to 8 inches where its visual size indicates
it should be. Very "natural" in its high depth... maybe the slightest touch
of "stretch." Maybe the taking/viewing factors in this situation "compensate"
each other enough for my brain to accept the result as near real.
-----

"It seems to me this is another argument for using reduced stereobase.
If you use a reduced stereobase, you increase all three dimensions
by the ratio of normal/used (that's 65 mm divided by whatever base
you actually used).  So if a macro is shot at a foot distance with
a 1/7th stereobase, the near point will be at 7' which matches
the nominal Realist near point.  Ought to work out well but of
course the reconstructed image will be 7X.  That's not entirely
bad of course but it's not entirely realistic either.  8-)"

-----
     The only reason I see for reduced stereobase is to reduce the amount of
image point deviation. If the scene image point deviations produced by the
"normal" stereobase are acceptable to you, don't reduce the stereobase. 
     This is the concept inherent in the Realist projection mask system. 
     The closeup and medium mask ranges, infinity not included, give about
the same image point deviations as the distant mask range of seven feet to
infinity. You are essentially "mounting to the window."
     What image point deviations are acceptable to you and how your brain
interprets the scene in the presence of variations from "ortho" in the taking
to viewing chain is a whole different discussion! :=)
-----

"An alternative would be to mask off the outer portions of a wide 
format mount (say an 8P RBT).  That would help with the vergence 
and window distance if the vergence control on the projector is left 
at its normal setting.  This is of course a good formula for ghosting."

-----
     I think you're describing the "double depth" mounting technique Fewerda
covers as a way to avoid window violations or far point(infinity)
divergence... way more than "double" if you are taking normal base at 12
inches.
-----

"The only remaining problem would be a problem for people who still 
have some accommodation and that is that focus distance is way off.  
I don't think this is too serious, though."

-----
     Double depth at least preserves convergence cues and scene objects at
screen distance also have correct accomodation, the Realist projection system
presents neither convergence or accomodation cues accurately, but the brain
gets it right even when the convergence from infinity up to the screen is
"filled" with just a few feet of closeup subject matter.

LeRoy Barco
LeRoyDDD@xxxxxxx


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