Re: Macro-stereobase

[bercov@xxxxxxxxxx (John Bercovitz)]

Jamie Drouin writes:

>      To figure out the 1:30 stereobase when using a single camera on a
> slidebar do you measure from the filmplane to the nearest subject or from
> the end of the lens to the nearest subject?  This becomes rather crucial
> when working extremely close and the two methods make quite a difference.

I'm 99% sure it's the primary nodal point you measure from.  There's a 
nagging thought that it's the entrance pupil (because it's the perspective 
point according to some) and I've never completely resolved that one.  I 
really must think on it some more.  But for on-film deviation, which is 
what you're talking about, there's little doubt in my mind, and it's worked 
for me in the limited number of macro shots I've taken.  

It's easy to find the location of the primary nodal point even if the name 
of it sounds scary.  Shall we call it the PNP?  8-)  Anyhow, let's say you're 
using a 50 mm focal length lens.  Take the lens off the camera or open the 
back of the camera and point the end of the lens that usually points at the 
film at an infinitely-distant object.  Now hold a white card in front of the 
end of the lens that normally points at an object.  Move the card until the 
distant object is in focus.  Measure 50 mm from the card toward the lens.  
That's the location of the PNP.  For the usual 50 mm lens, it will fall 
somewhere inside the lens.  It has a fixed location relative to the lens 
regardless of how you focus the lens (unless the lens is a front cell 
focusser).

I really wouldn't use the 1 in 30 rule for a macro; you'd be throwing away 
far too much depth.  Use either Mannle's approximation or Waack's 
approximation or the exact formula (the Bercovitz/Spicer formula) which 
isn't too much more difficult to use.  If you feel innumeracy coming on 
when you look at this sort of stuff, by all means send the particulars of 
the setup and I'll be happy to figure it for you.  Perhaps your example will 
make others more comfortable with using such formulae themselves.

Philosophically, if it's a macro that's not too tight, I would try to get as 
close to 65 mm separation as possible to avoid stereo scaling effects.  Of 
course if you want to scale the thing up, a perfectly reasonable option, 
that's another matter.  Just say how much and we'll see if it can be done.  I 
once scaled some "action figures" my son had painted up to 6' tall.  Worked 
well but with the limited depth in the scene, I probably could have left 
them at normal size without having to place them too far from the camera.  
Scaling is merely 65 mm divided by the stereo base of the shots.  So  if 
you shoot with a 130 base, stuff will appear half size and if you shoot 
with a 6.5 mm base, things will appear 10 times size.  The only limitation 
on this is that you can't violate your choice of the various 1 in 30 rules 
available.

John B


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