Easy vs. less easy. Heavy vs. light. Field of view vs. field of view. Turning a knob vs. rotating a camera.

Each of these two photographs is made from a few images. The top is made using three frames, the bottom used five. Stitched in Photoshop to make each one, using different methods. With one of it the camera didn’t move at all. For the other the camera had to be rotated between exposures. Very different cameras, optics too. A Canon 5DII with a 24mm shift optic and a Fuji X100S with the 28–ish wide angle aux. lens mounted in a panohead I had laying about. No nodal plate to correct for parallax error. In other words, not quite according to the pano manuals.

Canon 5DII, 24/3.5 shifted 12mm left and right, Photoshop stitched (reposition) and cropped to 2,35:1.

Fujifilm X100S with the WCL-X100 aux. lens, 5 images rotated at 15° intervals in a non–parallax corrected pano head. Photoshop stitched (circular) and cropped to 2,35:1.

The two show about the same horizontal field of view, but quite different on the vertical scale. Fundamentally different approaches to recording raw material as well as stitching, resulting in very different outcome.

As much as I’d love to love the 5D pano for its simplistic and orthodox approach (and the fact that I don’t have to purhcase anything to do it properly–ish) I seem to prefer the rotated results of the Fuji. And now I’m referring only to the geometry in the image, and how it lacks the stretched–edge effect that often shows up with rectlinear superwides. Of course, the stretched–edge could be compensated for, it might even be possible to record a simple action that automated much of the process. On the other hand I’m a hopeless Fuji fanboy when it comes to file quality, or character. Oh choices.

More testing is in order. I don’t quite understand this pano thing anyway.

Click pics for more pixels, at Flickr. Neither is checked for stitching errors.