4

When shooting a panorama, rotating the camera & lens about the nodal point of the lens is expected to produce higher-quality images (in the sense that they match better at the edges) -- hence the existence of panoramic heads and nodal sliders.

If I've got a nodal slider, though, I still need to find the nodal point of the lens. For primes, I understand that the nodal point will be fixed, so I could find the point once and mark it, but for zooms, since the nodal point moves when I zoom, a method that can be easily replicated in the field would be best.

How do I find this point?

1
  • 1
    FYI the correct term is either the centre of projection, optical centre or "no parallax point", the most widely used term: "nodal point" refers to something else!
    – Matt Grum
    Jan 24 '14 at 16:25
5

There is a nice video explaining how to do it here: https://www.youtube.com/watch?v=k0HaRZi-FWs. The procedure is the same with primes and zoom lenses and it works in the field as long as you have two vertical reference lines to work with.

I'm not really sure what method you're looking for for the zooms though. You can always set the lens to a certain focal length in the field and then calibrate it. If you want something quicker you simply have to calibrate it for some different focal lengths that you think you will be using and mark them. The movement of the nodal point (as a function of the present focal length) of the zoom lens does not have to be linear. It's not even certain that it will move at all, even though this is unlikely.

In short, perform some measurements at home and mark up the different focal lengths you think you will be using. If you in the field realise that you need another focal length you could either calibrate it in the field as shown in the video or perform some crude interpolation from the data you've already got.

God luck!

2

Panorama's.

Centre Nodal Point.

I wrote this for myself sometime back when I was considering buying a second quick release plate for my arca swiss style tripod which led me to investigate if doing so could improve the parallax errors in the panorama photo's I sometimes stitch together using Hugin open source software.

fig. describing terminology

The Centre Nodal Point (C.N.D.) is the point on the lens about which it must be rotated in order to eliminate parallax when joining images to make a panorama. It is often around the middle of the lens and varies with the focal length of a zoom lens (although - "FYI the correct term is either the centre of projection, optical centre or "no parallax point", the most widely used term: "nodal point" refers to something else! – Matt Grum" in the questions comments disputes this term).

The tables below were taken from www.olypedia.de. They show the distance from the sensor plane to the C.N.D. and the adjustment required for the distance from the sensor plane (θ) to the tripod mounting screw hole.

 E-1      -19mm
 E-5      +4mm
 E-520    +1mm

Table 1: sensor plane adjustment.

               to sensor   mount (+1mm)  to sensor   mount (+1mm)
 11mm          113mm       114mm         -              -
 14mm          108mm       109mm         76mm        77mm
 18mm          103mm       103mm         74mm        75mm
 22mm           98mm        99mm            -           -
 25mm              -           -         74mm        74mm
 35mm              -           -         70mm        70mm
 42mm              -           -         56mm        56mm

Table 2: distance from C.N.D. to sensor plane and mount.

These figures correspond roughly with those from www.myfourthirds.com by Klaus Schraeder (this link appears to be broken now) for the 11-22mm lens on an E-1 camera when adjusted for the sensor plane distance.

              to tripod mount  for E-1 mount  to sensor plane
 11mm         95mm             19mm           114mm
 14mm         90mm             19mm           109mm
 18mm         85mm             19mm           104mm
 22mm         80mm             19mm            99mm

Table 3: distance from C.N.D. to sensor plane and mount for E-1 after Klaus Schraeder

fig. showing camera panning across two targets

To find the sweet spot for a lens set the camera on a levelled tripod focusing the camera at two vertical targets separated by as large a distance as possible. When panning the camera from side to side you will see the targets move in relation to each other and the background if the lens is not properly mounted. Move the lens forward and back on the mount until the least amount of parallax is evident in the viewfinder.

two photo's illustrating parallax

These images above demonstrate to parallax effect created by panning the lens across a scene with one nearby object (the rock) appearing to move in relation to a distant one (the tree line), and below the effect after the camera's position was corrected.

two photo's illustrating corrected parallax

www.manfrotto.com suggested (I can't find the article anymore) starting with the front lens element over the tripod mount axis and moving the lens forward until the parallax error is eliminated.

I found I could use a flash bracket mounted on top of my existing 50mm quick release plate to gain the necessary offset, though the length of the bracket in front of the lens meant it entered into the frame on wide angle shots. By trial and error I found that the OM 50mm lens on an E-520 sits with the front of the lens roughly over the centre of the tripod head and the 11-22mm sits near the front of the focus ring when at 22mm.

If you want more parallax diagrams try this answer by rafael.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.