Graphics Doom Textures LoRA (Stable diffusion)

[smoke_th]

Alright, so. I trained 3 LoRAs based on doom and freedoom textures: doom-archival-v1, doom-v1 and freedoom-v1

https://civitai.com/models/156261

For basic explanations of what is what you can check here:

Spoiler

https://stable-diffusion-art.com/beginners-guide/

https://github.com/civitai/civitai/wiki/How-to-use-models

https://github.com/civitai/civitai/wiki/How-to-use-models#lora

Some examples (gens use revAnimated as base model, images are gen + nearest neighbor downscaled down to 128x128 and upscaled back side to side comparisons):

Spoiler

 

 

 

(also check image posts attached at civitai page, richyrich515 did a decent comparison breakdown with better prompts)

 

The purpose of this is obviously not for direct use, because textures come out mostly blurry, but instead to permutate and downscale them, author via pulling the elements, mixing and matching and probably applying palette. Keep in mind that only doom-archival-v1 has switch image data in it, other two were authored more thoroughly with flats being tiled 4 times and 64x128 walls being tiled twice horizontally, so that after upscale - whole dataset was 512x512 images, for sake of training. 

 

As for usage rights - the freedoom LoRA can be used to generate commercial images, just don't forget to chuck BSD license alongside whatever you using them in, as per terms and conditions of freedoom's BSD license. Other two are strictly noncommercial, for sake of obeying idsoftware licenses.

 

Now, with some grunt you can get unlimited textures. 😁

Edited October 4, 2023 by smoke_th

[pbuyle]

I missed a bit with your model and ended up with this creepy texture...

 

I posted a few more at https://civitai.com/posts/826716