The reason I use the Gimp is not because it's basic, but because it's by a wide margin the most sophisticated graphics software available in the public domain. The trouble with software correction of optical errors is that it involves interpolation. The software basically moves pixels to the places it believes those pixels should be. That's fine when your field of view is measure in meters, but we measure FoV in millimeters and, by necessity, individual pixel placement is vital to what we do. Interpolation doesn't work so well at that level; it's why, when reducing size, we try to stay within mathematical multiples to reduce the amount of interpolation required. The Gimp has_no_clue where the pixels *should* be. Perspective interpolation wouldn't work if you wanted to study individual bricks on a building, either. I'm aware the Gimp is capable of correcting perspective. Never used it, never will. Given the importance of pixel-level accuracy in our discipline, it's more efficient to use technique which makes such postprocessing unnecessary, and I don't trust the results when applied to someone else's work.
