Hi, in terms of showing diffs between two images, does this seem like a good way to show what's going on?
[image: image.png] Render pixels in both images "normally" (ie the intersection) Render what _should_ be there in green (ie baseline - intersection) Render what _shouldn't_ be there in red (ie test - intersection) I see the current system takes a different approach with blurring, but I figured given I have my hands in there, I might as well see if other ways to represent this same information work better for people. I seem to vaguely recall several years back the diffs were shown in roughly this representation. Another example: [image: image.png] Cheers, L -- Luca Fascione
