Archived posting to the Leica Users Group, 1998/02/06

[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]

Subject: Re: [Leica] Re: Distortion
From: dannyg1 <dannyg1@IDT.NET>
Date: Sat, 07 Feb 1998 03:42:43 +0000

Jacques,

> Consider any ray of light passing through the nodal point of a symmetrical
> lens (ie the middle of the diaphragm). Any bending of the light done by the
> lens elements on the way to the nodal point will be undone by the rear lens
> elements

(FYI nodal front will be a third behind the front element while nodal 2 will be a 
third inside the rear set.)

To continue your thought, as it's written above: 

And since rays bound for the image edge will have to travel farther, to their focal 
point, they'll also have longer measured focal length (point to point), increasing 
their effective image magnification at the their focussed imaging distance. 

A proof that this happens is the cos4 law, as well as the fact that extreme off-axis 
image distortion is present to any lens, symmet (close to the perfect lens) or not.

My assertion is that this gradual image magnifiction is perfectly sympathetic to 
countering central point perspective diminishment across the image field, which 
would produce pronounced barrel distortion in a normal focal length lens and 
fisheye effect in a wide angle, if left uncorrected while imaged to a flat surface.

I think the answers might be found in the imaging characteristics of retrofocus 
wide angles. If they project the light to the rear nodal point, and then refract the 
light obliquely (and very sharply), in air, _behind the lens, then (and only then) 
could it be possible that their oblique projection angle be equivalent to a symmet. 
Every ray trace I've seen (on retrofocus WA's) shows image cones being projected 
from lens rear in a cone shape and not, a trumpets bell shape.

Take a look at this URL: http://focus-software.com/zemax/index.html. In it, you'll 
see a ray trace for what looks to be the older H'blad 50mm. Notice that the raytrace 
shows no sharp deviation at (what should be) 50mm's from film plane. Also notice 
the narrowing of the oblique ray angle spread, as compared to the central trace. To 
me, that signifies a tighter effective focal length.

The Zemax site has an optical design demo that'll run on NT or '95. IMO, this URL 
will interest more than half of the list and stands as one of my favorite, interesting 
links yet found. All you armchair lens designers with Windows machines: Don't 
miss this one!

As I said, I need some answers that aren't readily available. Hopefully it won't take 
forever to get them. 

Regards,
Danny Gonzalez