A while ago I acquired a FA31Ltd that I have been trying to use for multiple frame panoramas stitched together. Initially, I got carried away with the speed of the lens and my DoF paid the price (along with a soft focus) - and I was not happy with the results.
So, I finally had some additional time this evening and was able to get some shots. I really missed last night's absolutely specular sunset, so I made due with this evenings. The result is the sharp images I was expecting - however, I did forget to reset the camera to ISO 100 and left it in automatic, so these are at ISO 400. The pano is 9 frames across or 16K by 4.5K, f5.6. f5.6 is the sweet spot for the 31, along with minimum vignetting and CA.
As a comparison I am posting 2 additional images of the same shot. The first one is from a couple of weeks ago, using the 12-24 at 24, f8 - 6 frames. The second is a shot with the 10-17FE at 10mm, f8 a bit too late to get the sky color I was after, but you can see that each of them are 180 degrees wide.
I also used a different stitcher that did a wonderful job in blending the sky across the stitched frames. Did much better than Autostitch.
Also, there is about 10 to 15 minutes between the pano using the FA31 and the 10-17, so I lost the really blue iridescent sky that I was after...
The 9 frame pano using the FA31 takes up the full width of a 23" flat panel display and its detail is wonderful. However, when you take all 3 of these individual images together, the focal length does play a critical role, especially when you scale each of these images to only 1000 pixels wide. The first one at 9 panels, really does reduce the height of the resulting image - that just comes with the territory. The second, at 6 frames, does not squish (an engineering term) the image as much. The third image using the FE - as a single frame really does frame the foreground a bit better than the first 2. Actually, sitting here looking at all 3 of them stacked up at 1K pixel wide really formed this composition observation.
I am finding that if you do not have the head on the tripod level, as you pan around, you get a spherical effect and thus not a level horizon. The way to correct for this is a setup similar to the Nodal Ninga, where the pivot is occurring in about the nodal point of the lens, rather than at the sensor's plane - where the camera sits on the tripod (One more toy for Santa to bring some future Christmas).
One last observation. There is a little slop on the width of the images. I did not crop them to be exactly the same view. I am thinking that I should probably do this to align with the 10-17 and maybe recover about 5% or so. I really was not expecting the focal length to play such a crucial role in the formatting and composure - however it makes perfect sense.
Last edited by interested_observer; 12-14-2009 at 10:22 PM.