Are Zoom Lens "No Parallax Points" a Linear Function of the Focal Length!

[dwswager]

I recently got my nodal slide and panning clamp and decided to practice finding the No Parallax Point using the only DX lens I own.  When I finished taking readings at about 10 locations as reported by the camera body, I fit an equation to it coming out as a 2nd order polynomial.  Except looking at the data and the resulting curve, the errors of the testing procedure probably caused the polynomial as opposed to a linear function.  This has implications for actually how much testing needs to be done and using any focal length on a zoom.

AF-S DX NIKKOR 18-200mm f/3.5-5.6G ED VR II 
No Parallax Point distance from clamp center:   y = -0.0002*(Focal Length)^2 - 1.0733*(Focal Length) + 111.09

[Ellis Vener]

Verry interesting but I  find real world testing to be the best way to find the offset distance: http://www.reallyrightstuff.com/s.nl/it.I/id.33/.f

[dwswager]

Verry interesting but I  find real world testing to be the best way to find the offset distance: http://www.reallyrightstuff.com/s.nl/it.I/id.33/.f

Yeah, I have RRS gear and that is the procedure I used.  But the implications, if true are significant.  A person would still want to test their sample, but could do maybe 3 focal lengths and be done.  Also, because a camera reports actual focal length, you can dial up anything you want and then use the calculated NPP.

[NancyP]

In the real world, one is going to test at just a few FLs along the zoom range. Nice to know that it is adequate.

[allegretto]

while one may be able to derive a liar or poly equation for a given lens... it falls apart with each design and must be empirically re-derived. Further, one doesn't need exact fl's for pano

So I don't see the value of derivation beyond intellectual fun

[Bart_van_der_Wolf]

In the real world, one is going to test at just a few FLs along the zoom range. Nice to know that it is adequate.

Hi,

That would depend on the optical construction though. Zoom lenses with internal focusing and moving groups of elements may cause a displacement of the entrance pupil that is non-linear (may even oscillate). I recommend to first check (visually) by looking through the lens from the front, and judge if the entrance pupil shifts visibly/linearly when zooming in or out. It's best done by also closing the aperture a bit to better visualize the apparent entrance pupil position.

Cheers,
Bart

[fdisilvestro]

Hi,

That would depend on the optical construction though. Zoom lenses with internal focusing and moving groups of elements may cause a displacement of the entrance pupil that is non-linear (may even oscillate). I recommend to first check (visually) by looking through the lens from the front, and judge if the entrance pupil shifts visibly/linearly when zooming in or out. It's best done by also closing the aperture a bit to better visualize the apparent entrance pupil position.

Cheers,
Bart

The entrance pupil position will also vary by changing the focusing distance

[Bart_van_der_Wolf]

The entrance pupil position will also vary by changing the focusing distance

Hi Frank,

In general that's correct, which is why I tend to calibrate for specific scenario's (e.g. landscapes focused on 3 metres, longer distances will not exhibit much parallax anyway, or close-ups, or telelens/'gigapano' work).

However, also here, in particular internal focusing (lensbarrel keeps the same length) may leave the more frontal lens elements (and thus entrance pupil) relatively stationary, while shifting the rear lens elements and/or groups fore/aft.
One just needs to test for particular lens design.

Cheers,
Bart