Ha, ha Foxy,
Normal rules to getting new kit seem to be applying. Either diabolical featureless grey skies or loads of rain, so a play seems a bit off at the moment.
In fact so dire is it today, I've taken time to do a pic and a couple of grahics to hopefully better explain what this is all about. I'm quite sure those with fancy pano programs are probably thinking I have gone a bit over the top with this
. I am restricted to manually doing the stitch in CS2, so for me it helps to eliminate as many variables, like parallax error, as possible before taking the series to be stitched.
The first graphic illustrates my starting point. I figured if you are going to do a pano, then a single strip done with the camera in portrait orientation is easier to manually stich than a two row job done in landscape format. As can be seen from the graphic the D100 has quite an overhang from the tripod centre in portrait orientation, but even though the wooden bridge was over a hundred yards away I still had alignment problems through parallax error. This probably would not have been so obvious in landscape as the parallax error would have been seriously reduced.
The second pic better shows the location of the lens centre and its relationship to the centre of rotation. This obviously becomes critical if near objects are to be included and less so for distant objects. Out of interest I marked the centre of the camera tripod mount and the sensor plane.
As a change from post#1 I have added the sliding plate attatchment as I have a spare plate for that, so I can easily change the 501 head between applications without the need to unscrew everything.
The next graphic is my easy method of lens centering. Yes I fell over articles on Nodal Points, lots of tech talk and formulas that just made my brain hurt. So I did it the nice easy way as in the graphic. Tripod/camera set up in the workshop and six feet away I placed a table with two 10mm bar ends spaced 30 inches apart. The camera was lined up so the bars lined up. When the camera is forward of the lens centre and it is rotated, the further bar moved left or right. The further forward of the lens centre the greater the movement of the further bar. Once lens centre is achieved the rear object does not move relative to the front object. Two sharpened pencils would work just as well for this.
I intend using two primes, a 20mm and a 35mm and luckily they both centre within 1mm, not too worried about that and I only marked one index point on the sliding plate. Proof of the pudding was when I found this site which lists a number of lens centres, and the 35mm is spot on with that, measuring lens centre to sensor plane.
http://www.swissarmyfork.com/lens_table_1.htm
For pano you need to look at the column marked ' Entrance Pupil '. This link also illustrates that with zooms that centre point moves with change of focal length.
Don