Artcameraman wrote:
It's more a case of the resolving power of the lens. True if you don't have to move it, the BL scanning back, vs. the 35mm camera which has to be moved 8 times, 4 on the top and 4 on the bottom, 7 moves after the first placement the less distortion not to mention you don't have to consider the nodal point.
I'm curious about how you get only 8 positions?
It seems to me that would cover (at best) only 3" x 4", or a panoramic format of 2" x 6"--which would be reduced to 5" by the camera design.
With a sliding back on a 4x5 view camera, using my Sony in pano mode, I'm expecting to get three strips: top, middle and bottom, all are 1.5" x 5", and there is a 1/4' overlap, which should make stitching easier.
With individual snaps with a digital camera oriented vertically, there would be three rows with 6 snaps on each row to cover the 5" width, and a small overlap. total 18 images to merge.
If the digital camera is horizontal I'd get 4 snaps to a row with about 1/3" overlap, and need a minimum of 4 rows with no significant overlap to get close to 4" coverage (16 snaps) but more likely 5 rows (20 snaps) to get good overlap and possibly a bit larger than 4x5.
I'm not worrying about the nodal point of the lens. I'm thinking that stays the same whether the camera has a single large sheet of film, multiple small pieces, or an electronic sensor moving around taking samples from different locations.
What I am concerned about is the possibility of vignetting at the edges of the frame when using wideangle lenses.
The mount for the digital camera will probably get in the way of light coming from an extreme angle. A flatbed scanner would give the best results in the edges and corners.