You know what I've never seen? A crop from a well-exposed original image that needed an exposure boost per the offered theory. So I decided to test said theory with a real life example...
Primary image is a FF generic image taken with the A7R6 and 24-70 GMii lens, processed basically as-shot SOOC. Showing full frame, then a crop to 25% of the image area from the center; and finally that same crop with the suggested 2-stop exposure increase to make the FF and 2x crop exposures "equivalent" per the stated physics...
Notes: In the full image the darkest pixel I could took the R channel down to 0 with readable values on the G and B; in the brightest white area, the hottest pixel I could find was B at 252 with G and B something less. In the zero change crop, the dark pixel still measured the same, but the hottest pixel got cropped out and now the hottest pixel was B at 247. This all before conversion to sRGB for this web view. After the 2-stop exposure boost, the R pixel stayed 0, but the G and B increased notably, and of course there became a whole lot of pixels, basically the entire sky with B fully blown at 255.
PS: And FWIW I could not detect a single iota of increased noise in any shadow region after the crop. I am not claiming it wasn't there, in fact I am certain it would be if I had a scientific way to measure it. I'm just claiming I could not visibly detect even a hint of noise anywhere even at 400% magnification as a practical point.
Now it's my turn to run for cover
Sep 21, 2026 at 10:25 AM
Previous versions of Jack Flesher's message #17110434 « New Lenses? »
