A dim photo can lose visible detail during compression even if the highlights look fine. Dark gradients and textured shadows often show breaks, color noise, or smeared borders before anything changes in bright areas. Choose a reference region and review the compressed output at the brightness the audience is likely to use.
Choose two shadow areas
For a photo of a black bicycle inside a dark hallway, pick a section of the frame where the wall should remain smooth and a section where the tire tread has real detail. Compare both at normal display size. If the image relies on the tread to communicate condition, that region has a firmer quality requirement than the smooth wall. Mark the lowest setting that preserves the important shape in a comparison copy.
Lower bytes without making the photo darker
Try a clean crop or a smaller pixel target before resorting to very strong compression. Save two candidates at the same dimensions and inspect the actual shadow pixels in the final layout. If a few weak artifacts are visible only when enlarged beyond the destination, you may have an acceptable result. If dark areas merge into a solid mass at normal size, move to a better setting.
Review the delivered file
Use the image compressor and find the lightest compression that fits the upload budget. Reopen the result on a second screen if safe shadow separation matters. Confirm that the important shape is still recognizable and the photo keeps its intended atmosphere. Keep the original for later displays where the byte cap is different.


