T3D Re: Focal length

[john bercovitz <bercov@xxxxxxxxxxx>]

Tom Deering writes:

> I'd like to continue to use John's formula, but I want to make sure I
> understand the concept of focal length.  If I place a 45mm lens on a 
> 100mm bellows, do I now have a 145mm lens?   Is there a different way 
> to calculate focal length?  

Well, I somewhat tongue-in-cheek refer to the concept as the "operating
focal length".  That phrase would horrify any optical engineer because
there is actually only one, fixed, focal length for a (non-zoom) lens.  
To really clarify what's going on, I think the first thing to put at your 
T3D website is a paper on perspective (it's in Adobe Acrobat).

To answer your question, the maofd spreadsheet will calculate the 145 mm
from the 45 mm and the distance to the subject.  It will use the 145 mm
after that.

Maofd assumes you will then view the image with a 145 mm lens or sit the
equivalent distance from the screen when viewing a projected image.  If
you don't do this, you won't have an ortho image and I make no guarantees.
That's why they call me "Orthoman".  8-)  I always try to have the same 
viewing and taking perspectives unless there is some desired special effect.

> And does the formula work at such distances, or does it break down for 
> some reason?  I haven't received the bellows yet, so I can't run tests 
> of my own.  Yet!

This formula is the only one that doesn't break down at these distances
because it makes no simplifying assumptions.  (The usual formulas can
all be derived from maofd by making one simplifying assumption or another.
See P3D archives where various people have proven it for the popular 
formulae.)

Of course, you first have to agree with the underlying concept of maofd which 
is that photos with 1.2 mm of on-film deviation are fusible if the system
focal length is 35 mm and 2.7 mm is OK for a system focal length of 80 mm
and so on - in other words the on-film deviation shouldn't exceed 1/30 of 
the "focal length" of the system for good fusion.  As noted in an earlier post
by Jim C, Boris comments that the rule is mutable according to the magnitude 
of the disparity gradient.  I agree completely.  I think you can go to 1 in 15
with a very gentle gradient because I have done it, though not in a macro
situation.  1 in 15 is pushing it, though, I have to warn you.

John B


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