More on the pano pivot point

In the previous post I discussed some well known ways to find the location of the front entrance pupil, eg using the 'pole technique', using a pencil light beam or using some lens modelling and simple maths.

Unless you are interested in the details of lens design, and most photographers are not, then all you are interested in is the answer to the exam question, ie: where do I put my camera so I can achieve the best pano stiching, ie minimise the pano parallax artifacts.

Finding the 'pano pivot point', ie the entrance pupil or no parallax location is, as we have seen, easily achieved in the field (assuming you have some near and far reference feature to use, eg the 'two poles' technique).

However, if you are going to find this point for all lenses, then you need to open your mind to what's going on, as 'seeking' this point should not be limited to within the lens or camera body.

As an example, let's take the Canon 24-105mm F/4L lens, in this instance on my 5D3. The lens layout is as follows:

A quick way of estimating the location of the pano pivot point (plane E in the DOFIS model), is to shine a pencil beam into the camera. 

As mentioned in the last post, if the light is aligned correctly, you will see a maximum blooming on the live view. These two screen captures show what you are looking for. In the first image the light is not aligned; in the second it is:

Looking from above, as in the last post, one can estimate the entrance pupil location, which, at 24mm and focused at infinity, correlates 'perfectly' with the DOFIS feedback, ie 134mm from the sensor plane. That is the entrance pupil is where you might expect it to be, and thus the nodal rail can be positioned accordingly.

Let's now look at the 24-105mm lens at the other zoom extreme: 105mm. Where do we think the entrance pupil is now? 

Unless you are an optical lens wizard, you will be hard pressed to 'guess' this location. Although the focal length is likely to give you a hint, ie a zoom with a long focal length.

So let's shine a light into the lens and see what happens:

So there you have it: the entrance pupil, ie the pano pivot point, is positioned over 100mm behind the camera, based on a very quick assessment. In fact, looking at the PhotonToPhotos optical bench, the infinity focused entrance pupil is some 77mm in front of the sensor.

So a word of caution is needed at this point. The above test is not a very accurate way of doing this technique, ie simply placing the light on a surface and pivoting it. To really use the technique you should ideally use a low powered laser, hold the laser in a clamp and measure the angles more accurately than I have here. Finally, I would always do a second estimate using the parallax method, and, if possible, see if the lens is in a data base, such as PhotonsToPhotos (see link on right). 

Ok, we know where the entrance pupil is, but is it useful to know this? Maybe, but to be fair, most pano captures with a near field object will be with a wide or normal focal length lens, or if shooting with a tele, there won’t be an object too close. But at least you now know.

Bottom line: if you are a pano shooter, you may wish to 'get to know' your lens, especially a zoom lens, by carrying out a simple torch test as above. This way, you will have a fighting chance to fit your nodal rail the right way round!