Doug Kerr
Well-known member
We continue to hear (thankfully not as frequently as a few years ago) about the need, in multi-shot panoramic photography, to pivot the camera about an axis passing through "the nodal point" of the lens.
Those even slightly familiar with photographic optics should be immediately suspicious of this phrase, if only from the standpoint that lenses ordinarily have two nodal points. Through which one should the pivot axis pass?
The objective here is to rotate the camera about an axis such that there is not a parallax shift between successive shots. Such a parallax shift interferes with the proper joining of the separate shots.
Some authors refer to the proper location of the pivot axis as the "non-parallax point", which is certainly a good functional description of it.
And where on the lens is that? Well, it can be shown that the pivot axis should pass through the center of the entrance pupil of the lens. The entrance pupil is what the aperture stop looks like, and where it looks like it is located, seen from in front of the entire lens. Might it coincide with one of the nodal points? Yes. It might coincide with front of the focus ring. But either of those would be essentially accidental.
Some who are anxious to continue to believe that "the nodal point" is the proper location for multi-shot panoramic work say, "But I used the test to find the nodal point, and set my axis there, and have no parallax." I ask, "Which test is that?" They say, "You know, the one in which you vary the pivot axis until you see no parallax shift in the viewfinder as you rotate the camera about that axis."
That is indeed, by definition, the test to locate the "non-parallax point". But it is not the test to find either nodal point.
The test to find the rear nodal point involves finding the axis such that, when the lens is rotated about it, there is no image shift on a fixed ground glass, quite a different matter. But it might be the superficial similarity of the two tests that is responsible for the misunderstanding.
Best regards,
Doug
Those even slightly familiar with photographic optics should be immediately suspicious of this phrase, if only from the standpoint that lenses ordinarily have two nodal points. Through which one should the pivot axis pass?
The objective here is to rotate the camera about an axis such that there is not a parallax shift between successive shots. Such a parallax shift interferes with the proper joining of the separate shots.
Some authors refer to the proper location of the pivot axis as the "non-parallax point", which is certainly a good functional description of it.
And where on the lens is that? Well, it can be shown that the pivot axis should pass through the center of the entrance pupil of the lens. The entrance pupil is what the aperture stop looks like, and where it looks like it is located, seen from in front of the entire lens. Might it coincide with one of the nodal points? Yes. It might coincide with front of the focus ring. But either of those would be essentially accidental.
Some who are anxious to continue to believe that "the nodal point" is the proper location for multi-shot panoramic work say, "But I used the test to find the nodal point, and set my axis there, and have no parallax." I ask, "Which test is that?" They say, "You know, the one in which you vary the pivot axis until you see no parallax shift in the viewfinder as you rotate the camera about that axis."
That is indeed, by definition, the test to locate the "non-parallax point". But it is not the test to find either nodal point.
The test to find the rear nodal point involves finding the axis such that, when the lens is rotated about it, there is no image shift on a fixed ground glass, quite a different matter. But it might be the superficial similarity of the two tests that is responsible for the misunderstanding.
Best regards,
Doug