!!!This is AI generated!!!

So new workflow and way of donig things, i would like to think that my eyes and hands got a lot better but of course it's still a mixed bag

The first 8 are with afterwork done by me with inpaint and Photoshop
from there until 18 is simply stuff that i created with my basic setting and new workflow
from there is my venture into more detailed background and "themes"

!!!Remember AI GENERATED!!!!

Bimbo is awsome

Base +9, Nystic +6, prokeds +6, xhoextms +6, MehMeher -5, glitcher998 +9, MCHamon +6, angrytaco -6

dammm yes,,, delicius girl.. omg .. i remenber mi exgf Yoselyn

Base +5, prokeds +6, MehMeher -5, angrytaco -6

it looks so delicious

Base +1, xhoextms +6, MehMeher -5, prokeds +6, angrytaco -6

A pretty significant step up from your previous set, you're doing very well!

Base +6, MehMeher -5, prokeds +6, angrytaco -6

Thanks a lot, and i think the step up to a better outfit and background selection is also in full swing

Base +1, MehMeher -5, prokeds +6, angrytaco -6

Make some more with bimbo concept. and please make a pixiv account to follow.

Base +11, angrytaco -6

I have a Discord where i post most of my stuff https://discord.gg/RuQ52ABV <- if its not valid anymore write me a PM i'll try and answer or update

Last edited on 22 April 2023, 07:33.

Base +1, angrytaco -6

Amazing stuff, great job!
Which AI tool(s) are you using? Besides your afterwork that may be quite some effort!

Base +6, angrytaco -6

@booblover1
>Which AI tool(s) are you using? Besides your afterwork that may be quite some effort!

I Use Stable Diffusion on my local rig with some basic add ons.

And currently it's nearly no effort to generate consistently pics that are at least good, since i started the whole project with automation in mind. I wrote a small prompt script which generates neraly everything automatically for me (pose/cloths/colors/themes etc) and what i mainly do is ironing out the overall problems like the eyes(i.e. searched/made/found loras and TI)/hands/general sharpness and composition as well as adding more models/data into the mix and refining the script (last project was a bigger lips prompt that i could throw into the mix without fucking things up and next will probably the hands). New outfits and themes and poses are also quite time consuming since i want to make sure that the new prompt does what it should and that requires testing. Also generating cohesive sets like the bar in the thierd gallery requires extra work

The after work for the pictures did mainly consist of removing small extra stuff like a small droplet on the skin or a small color dots, and that problem i could already solve of. So not really that much that required change they pretty much come out like that..
And i think i'm currently at a point where the only real problem are sometimes the hands/fingers and occasionally a bad iris, but overall i would say when i start my dynamic Prompt "script" i get around 7-8 good pics from a batch of 10 . Still with small errorrs but they could probably be corrected in 5-10m photoshop, and some come out really good i.e.: https://imgur.com/a/AAYecPf all of them came out like that.
Still with really small problems like the earing on the first and the fingernails on the second or the ending of the left glove in the last pic but that's a level of errors i would be fine with. I would die happy if every pic came out like that :D

Last edited on 15 April 2023, 01:19.

Base +1, angrytaco -6