[Pigking/CrazyDad] Comics

So here we are: Emily 6 preview.

Page_01.jpgPage_02.jpgPage_03.jpgPage_04.jpgPage_05.jpgPage_06.jpgPage_07.jpgPage_08.jpgPage_09.jpgPage_10.jpgPage_11.jpgPage_12.jpgPage_13.jpgPage_14.jpgPage_15.jpgPage_16.jpg
 
I'm testing the AI for another reason: to remaster comics, especially older ones. Here are some examples, and you can decide which one you like best. For me, ChatGPT and Grok give better results, but they have a very strict filter on sex scenes. I couldn't even get Microsoft Copilot to work. If you know of another site or program to improve images with AI engines, please let me know.


ORIGINAL


LF1 (12).jpg






A2E

a2e.jpg





CHATGPT

chatgpt.png





GEMINI

Gemini.jpg







GROK

grok.jpg








IMAGINEX

imaginex.jpeg








TOPAZ

topaz.jpg
 
Nice, thanks. I scaled them all down to the smallest (grok, 1168x880), added the AI name and cut the images to the womans head:

LF1-a2e-part.png
LF1-chatgpt-part.png
LF1-gemini-part.png
LF1-grok-part.png
LF1-imaginex-part.png
LF1-orig-part.png
LF1-topaz-part.png


Although they're the two best, I'd say chatgpt and grok are out of the race because of their filters.

From the remaining a2e comes last because of its blurred textures, gemini isn't that bad but the image size is too small, but topaz seems quite ok and can output the original image dimensions, can it?

If yes, please try a topaz run through something more explicit like LF4, if possible.
 
Onomatopoeic words help with the context by reinforcing the action in a composition. However, covering the artwork with them causes a lot of the erotic atmosphere to be lost. Just notice the difference by simply changing the position of the words.
Sem título.png
59.png
 
Back
Top