If I had a Q3 43, I could well want to crop to simulate the FOV of a 100mm lens. This would result in a significant drop in the number of pixels. Has anyone done similar and then used LR super resolution to up the number of pixels 4x.
Lightroom's
Super Resolution is still limited to 2X, but you can use Topaz
Gigapixel within Lightroom for a 4X (or greater) enlargement. (Adobe recently acquired Topaz; apparently, the plan is to keep the brands separate, at least for now, despite their common ownership.)
In my experience, Adobe's approach to AI enlargement is conservative: SR seems more likely to extrapolate additional pixels rather than using a neural network to generate them. Having said that, I occasionally do see generative-AI artifacts with both enlargement tools.
Gigapixel tends to be more aggressive, although some of its models offer a slider you can use to throttle the extent to which it invents data that do not exist in the original image. There's no way to determine in advance what the neural network will do with a particular image (and, in fact, it will perform differently each time you use it with the same image), so for both a 2X comparison with SR or for a greater enlargement, the best approach seems to be to try them both and decide for yourself which result you like better. Sort of like using Perceptual and Relative Intent when you're soft-proofing for a print.
