I see one problem, but this may be more up to the examples.
look at the grey one here >>626
it may be easier to see.
there are subtle differences in the compression
if the images were exactly the same, the image would just be gray.
The example isn't the best as these are both scaled down from the 'master' size, but it shows what happens in a difference when the images are more or less identical.
If imagemagick can do this without highlighting the entire image, then it may be worth pursuing, the main use I see from this is being able to see hard differences with the difference in your face, see the example I just made, sometimes it's not even dicks, it could just be down to one image was more finished and has let's say sweat in it instead of the lesser finished one, maybe the hair is a bit more done. this method would highlight everything that changed in a very visible manor,
while its also able to do it in a very subtle manor, see the grey megumin that was not only a different size, but if I remember right, a similar compression ratio, for a complete set look for the black ones I made in a prior version,
The concern I have with it, is partially the VERY small images they use as examples, the thing looks like a sledgehammer, never giving you all the details, just what it thinks you want to know.
Honestly, I think the program is closer to a surgical scalpel then imagemagick is, it's just a matter of overlaying a 50% opacity inverted image on top of another, granted me saying that is assuming its work, but there are functions you can call upon to do this that the dev never considered. Like I said before, the heavy lifting is more or less done as far as I can tell
Message too long. Click here to view full text.