Oh hey, a newbie just caught the AI image bug

[Sekani]

The Good One also is where most of the usable indoor public spaces are. Well, technically usable. Places like cafeterias and shower rooms no longer serve their intended purpose for obvious reasons, so this tends to be where the inmates gather to do... whatever is it they do that doesn't involve sex or violence (more on that later).

young_women_with_huge_breasts_ 9.jpeg

young_women_with_huge_breasts_ 66.jpeg

two_young_women_with_huge_brea 10.jpeg

young_women_with_huge_breasts_ 69.jpeg

[Viridian]

Welcome to the rabbit hole. Here's a taste of what you could unlock once you get more familiar with the tools.

[Sekani]

The Bad One is the older of the two buildings, and instead of private rooms still uses small metal-barred cells. The building's integrity has also greatly suffered because of the whole sinking into the ground thing. Usually if you end up here, you're either new or you pissed someone important off.

young_women_with_huge_breasts_ 31.jpeg

young_women_with_huge_breasts_ 10.jpeg

young_women_with_huge_breasts_ 19.jpeg

young_women_with_huge_breasts_ 51.jpeg

young_women_with_huge_breasts_ 25.jpeg

[Sekani]

Welcome to the rabbit hole. Here's a taste of what you could unlock once you get more familiar with the tools.

When I first started goofing off a week ago, I didn't see myself ever advancing to this level. Today, I've got Stable Diffusion running on my iPad. I've got other ideas in mind that Perchance wasn't able to handle, so we'll see how it goes!

[Sekani]

The Bad One lacks the open public spaces that The Good One has for multiple reasons, but the inmates find places to stretch their... arms where they can. They're not expressly forbidden from using the more spacious Good One hangouts, but getting all the way over there can be a bit of a slog.

attractive_young_women_with_hu 4.jpeg

young_women_with_huge_breasts_ 26.jpeg

young_women_with_huge_breasts_ 28.jpeg

two_young_women_with_huge_brea 17.jpeg

two_young_women_with_huge_brea 16.jpeg

[MadMax359]

my Fave is in the hallway, she needs all the space she can find

[Viridian]

Looks like you're going through the phase I went through months ago. It's initially empowering to realise you can generate an infinite amount of images with just a button press - which personally to me kills the excitement when you realise that it takes virtually no effect to put out a decent quality image once you have the formula.

The way forward is experiment with different models (SD1.5 or SDXL), possibly using LoRAs (though I currently don't use any) and figuring out what your specific AI style is. Most people who are experimenting with AI at the moment peak at what you're doing now, so it would be interesting to see how you put your twist into AI images.

The immediate thing I noticed was that you have the same problem that most AI images have (including when I started) - the characters have no expression and look bored out of their mind. This is a pretty hard limit for certain AI models, and it takes some discovery to find which models are better at capturing expression. This is the part of AI art that makes everything look uncanny.

The other thing is to find a model that is better at creating mud textures. Lots of AI models go between the extremes of water and vinyl.

[Sekani]

The way forward is experiment with different models (SD1.5 or SDXL), possibly using LoRAs (though I currently don't use any) and figuring out what your specific AI style is. Most people who are experimenting with AI at the moment peak at what you're doing now, so it would be interesting to see how you put your twist into AI images.

I've got a few more Perchance images from this current prison set that I'll be uploading, then moving on from there. What I want to create next seems to be beyond the limits of what that site can do.

Going from Perchance to Stable Diffusion is definitely like taking the training wheels off. Having to manually handle a lot of the prompt engineering that was previously done behind the scenes is definitely a thing. At the moment trying to decide which model I want to invest time into learning is my biggest issue. The SD 1.5 models are the only ones ready to generate content relevant to our interests out of the box, yet I'm entranced by the potential of XL and Flux assuming I can find the right LoRAs to turn off the built-in censor. Already signed up for CivitAI cause I figure I'll be spending a lot of time there.

[Sekani]

my Fave is in the hallway, she needs all the space she can find

Always nice to meet another enjoyer of the well-endowed

[Viridian]

The SD 1.5 models are the only ones ready to generate content relevant to our interests out of the box, yet I'm entranced by the potential of XL and Flux assuming I can find the right LoRAs to turn off the built-in censor. Already signed up for CivitAI cause I figure I'll be spending a lot of time there.

While the bulk of my work this year was done on 1.5, I currently use XL.