Made some Ranger Fanart

Incredible 9.2/10 for me , personally i wanna suggest Berserk if you wonder what to make next. Good Job.

Very Pog

"

impulze3 เขียน:

alo, this is my ranger meowmyr

Just WOW :) great work

Amazing, pure awesomeness

спасибо, подрочил

Why not mention that this is AI generated? Even if you did some touch-ups (I saw your reddit post with the original image), it just seems quite disingenuous.

Where is this reddit post?
Would like to check out the original vs this touched up one...

"

Daredust เขียน:

it just seems quite disingenuous.

why ? you think i just pushed a button and out came this image ?

"

BoboTheExile เขียน:

Where is this reddit post?
Would like to check out the original vs this touched up one...

this is the base image i started with after generating around 200 variants for the overall base composition


i had to redrew a huge chunk of the image and what stayed the same is essentially just the composition and color palette. this was done partly by hand, partly by inpainting and variation via stable-diffusion, which is available as plugin for adobe photoshop and available on the app i use, krita.

first thing i did was add the bow by drawing it, aswell as drawing new chest clothing to make it sfw, i remade the hair and parts of the clothing, moved the belly button and remade the chest multiple times, redrew both arms and the hand a bunch of times (i really suck at drawing hands), then using neural render to blend the hand with the rest of the image and iterate over it to try get fine detail variations (didn't work as neural renders really suck at this and they ruined the hand over and over)

until it looked something like this



there is a big difference to the image the AI spat out, mostly the composition and color palette stayed but the rest i had to draw over and redo multiple times, i really get annoyed when people say its disingenuous to not say "it was created by AI" because it wasn't "by ai" it was created "using ai", as in any other artwork people make nowadays because most tools are machine learning assisted, like denoising, blurring and many other basic adobe features people take for granted nowadays. i created this image and AI was used as tool in that process but its not the entire process. i spent many hours drawing on my tiny little wacom tablet.

anyways, these are the other iterations i made along the way, note that all of these are AI upscales which introduced some artifacts later on, like on the hand.

"

redid arms, belly, fixed clothing, added details like feathers, quiver, knife

"

changed breasts, clothing once again, added more detail and added a new ear

"

idk what i did here but some small detail stuff

"

changed some really fine detail stuff, skin texture, small scratchmarks, blushing, scars (thanks to nacho for the suggestion)

"

impulze3 เขียน:

why ? you think i just pushed a button and out came this image ?

Props for the detailed write-up. I totally underestimated the amount of work you did, so apologies for that. I only remembered the ear looking different in your reddit post, but I wasn't able to verify because I couldn't find it anymore.

Do you have other arts? Other gallery? I really like this picture, and i would like to have another format to put it as wallpaper on my screen :)