Can someone explain the image on the right? what it is showing/how it is showing it? I know that the circles on her arm/shoulder are where the editing happened but why is it circles with one having further apart circles than the other? also whats with the big mark all around the top half of the picture?
Someone plz explain I'm actually a bit curious. Is she like a doppler? A gremlin? The Grey thing is when you look at her through the spiderwick lens thingy? I need context
This is just a random Japanese girl who happened to post a pic of herself in her office uniform. And a bunch of people on Twitter found it, thought she was cute, and started comparing her to a photo of Sydney Sweeney from her recent American Eagle collab, saying they preferred the "random Japanese office lady" over the "slutty western celebrity".
It's basically a rehash of the "supermodel vs random McDonald's cashier" meme, but this time with an added element of weeb/passport bro "slutty, evil western girl vs pure, wholesome Japanese girl".
That's the background. Now, as for what's happening in the thread, people found out that the Japanese girl edited her picture. Someone posted a screenshot from some sort of photo analyzing software that I guess is supposed to prove whether a pic is edited or not. That's the weird-looking grey picture. And the user who posted it happens to be in India (indicated by the Indian flag). Someone noted that the analyzed pic looks like Sharingan vision from Naruto, but made it a pun with "saar" which is a racial epithet used against Indians making fun of the way they pronounce "sir". Thus "saaringan".
Not sure, but it looks like an exagerated contrast filter (i think there's a proper name for this technique, but i forget) where the white of the picture shows areas where 2 or more significantly different colors/shades meet and the black shows the opposite. You can see white lines where certain objects in the foreground meet with objects in the background.
This normally doesn't happen in photoshopped images, as editors don't normally put in the effort to hide every off-putting contrast and instead simply make the image look passable in human eyes. The end result is that the edited parts end up looking full of low-contrasting colors/shades (for a computer, at least).
So I made a program that does this as well, but mine isn’t nearly as good as this method. Is there a repo or site for this or anything? Mine is based off of the [ELA method] (https://29a.ch/photo-forensics/#forensic-magnifier), but this looks like noise analysis.
38 Comments
sorryiamnotoriginal@reddit
Pyrimo@reddit
Rustymetal14@reddit
GoodwillTrillWill@reddit
Dqueezy@reddit
Rustymetal14@reddit
Dqueezy@reddit
Appropriate_Time_774@reddit
Rustymetal14@reddit
Retro_05@reddit
futainflation@reddit
stillmahboi@reddit
M3taBuster@reddit
stillmahboi@reddit
HFentonMudd@reddit
Only_Champion_8895@reddit
M3taBuster@reddit
stop_talking_you@reddit
stillmahboi@reddit
JCpac@reddit
JCpac@reddit
WankerAuterist@reddit
Helpful-Swan394@reddit
ProFentanylActivist@reddit
maxedonia@reddit
philmarcracken@reddit
MrDagoth@reddit
ormagoisha@reddit
MajesticStallionJean@reddit
sgtjoe@reddit
isuxirl@reddit
gecked@reddit
Crazy_Crayfish_@reddit
MrDagoth@reddit
heapzz@reddit
PikachuKiiro@reddit
Scary_Nail_6033@reddit
AlwaysHungry815@reddit