Danbooru-based image board with a specialization in high-quality images.
This post belongs to a parent post.
bikini_armor devil shingeki_no_bahamut tachikawa_mushimaro wings

Edit | Respond

why so many versions of this o.O?
eccdbb said:
why so many versions of this o.O?
This were in discussion of being fixed for the best quality possible.
It seems there were errors in previous versions. The others may be deleted or be grouped as Parent and Child when a consensus has been reached.
what filter/settings were used on this?
EDIT:
I doubt by now, since it has been hours, you'll see this eccdbb but you can go here & here; read both comments for the full information so far leading to this...I assume.
Just basic set by midzki years ago with parameters tweaked. It's a long story though, go forum to find the details of image editing.
Is there a hair on her nose and left breast?
I see. Maybe that's why the other screening and overfiltered versions were kept, they don't show as much dust on them, i still like this one the most.
It doesn't really matter...the source is terrible and any amount of editing will end up looking just as bad unless someone is OCD enough to edit every individual speck of dust.

Wiping the the material before scanning it shouldn't even take longer than 60 seconds in the first place and would have eliminated at least 80% of that awful crap.
I've uploaded my own attempt at filtering Post #305115 from scratch with the goal of eliminating most of that dust.

The end result was Post #305540 which I've hidden from the index for the time being, until it's decided whether it should be kept or not. A minor variation that post with a simple "Levels - snap to neutral midtones" at the very end was uploaded as Post #305536. At least one of these two should definitely be deleted.
WtfCakes said:
It doesn't really matter...the source is terrible and any amount of editing will end up looking just as bad unless someone is OCD enough to edit every individual speck of dust.

Wiping the the material before scanning it shouldn't even take longer than 60 seconds in the first place and would have eliminated at least 80% of that awful crap.
Oh my.
So this one is just doomed to having those "errors" unless a different re-scanned source presents itself right?