Any plans to add a pixel-identical deduplication system? As in, if a duplicate is detected with the "outline fuzzy matching" system then render out the jpeg or bitmap or whatever the file is for each potential duplicate to an array of ARGB and diff it, and if there's no differences run the normal "duplicates of equal quality" operation?
I understand Hydrus can't compare two jpegs to see which one is better, and that if the files aren't exact matches then a human needs to be in the loop somewhere, but I keep running into duplicate photos which have been compressed or had metadata stripped vs the original while in the comparison tool and can't help but feel that comparing identical arrays of bytes should be the computer's job.