Tesssarboy,
Tessarboy wrote:
> Doug Kerr wrote:
>
>
> Allright, if the PIVOT AXIS for composed panoramic photograps is
> not the nodal point you described, where is the pivot axis we need
> then located ?

The pivot point we need passes through the entrance pupil. That is the virtual image, as "seen" from the front of the lens, of the aperture stop (iris or diaphragm).

Here's why this is the proper location of the pivot point. (Be patient!)

In our panoramic work, we want the pivot to be at the camera's "point of perspective", which we can also think of as its "vantage point"; it is the place where the camera would be if it were a tiny eyeball.

If we pivot about that location, the "eyeball" does not move from side to side. If it did, the relative positions of near and far objects would shift (parallax shift), and this would of course screw up the stitiching of the diffferent shots.

The entrance pupil is, in the frame of reference of tthe world outside the lens, the aperture through which all light must enter the camera system: accordingly, it is the "peephole" through which the camera sees the world; it is where the "eyeball" seems to be.

Imagine that, for some wierd reason, we bulit a lens in which the physical aperture stop was not in the middle of the lens but rather a little bit in front of the front element of the lens. In such a case, the physical aperture stop is itself the entrance pupil.

It should be obvious that the camera can only see the world through this orifice - everything else is blocked. It must be the "peephole", and it defines teh "vantage point". If we pivot the camera about an axis through this orifice, the "vantage point" will not shift from side to side. If we pivot about any other place, the "vantage point" would shift from side to side., causing parallax shift.

Of course in "ordinary" lenses, the entrance pupil is inside the lens. But that doesn't make the "demonstration" any less valid.

Now how can we determine where the entrance pupil is? Three ways I know of:

1. By calculation based on the location of the physical aperture stop and the locations and focal lengths of all the lens elements in front of the aperture stop. (You could do this if you had complete construction information for the lens.)

2. By empirical measurement, in the way that is mnost often described in connection with locating the panoramic pivot point. (These are the techniques that are often mis-described as "finding the nodal point".)

3. By looking into the lens from the front with a microscope having a calibrated standoff scale, and focusing on the iris (actually, on the virtual image of the iris, since we're looking through one or more lens elements). (This is in effect a short-range "laboratory rangefinder".)

If you want to play physically with the "thought experiment" I used above, punch a small hole (1/4" will probably work) in a friend's lens cap, put it on the lens, leave the "regular" aperture stop open all the way (so its gets out of the act), and make the classical panoramic pivot poiint determining test. You'll find that the pivot point is now right at the lens cap!

Is this stuff neat or what?

Best regards,

Doug