Everything sounds like good advice. What do you do when part of the pano includes shadow areas that you want to boost in the final stitch? Do you recommend adjusting exposure on the camera (using shutter speed, I would think) or in post processing.
I use exposure bracketing and first create tonemapped tiles from the bracketed shots, which I then stitch. Depending on the situation, you may even get away with variable exposure time single exposures, but only if you use liberal amounts of overlap (say 50% overlap between adjacent tiles). It does require a competent blending engine to equalize the resulting exposure gradient.
I do have the basic RRS pano setup, but would like to occasionally do 2 row panos without buying the complete RSS pano setup. I have thought of getting a simple tilt head (such as the Manfretto monopod head). Most of my work involves rather distant scenes, so parallax might not be a problem. Do you think that would work?
When you dedicate 1 row to the area that involves foreground detail, then you'll obviously get a pefect fit there (assuming rotation around the no-parallax axis). The other layer row will probably blend reasonably well especially if, again, a liberal amount of overlap is used.
PS what is Lancroz 256? A simple google search is not helpful.
A Lanczos 256 (16x16) pixel area kernel is indicated, also know as an 8-lobe support Lanczos filtered Sinc filter (+/- 8 lobes * +/- 8 lobes in orthogonal directions). While it allows accurate signal reconstruction/resampling when small sub-pixel displacements are required, it also is prone to creating ringing artifacts with a higher amplitude closer to edges.