metadata

license: creativeml-openrail-m
language:
  - en
tags:
  - stable-diffusion
  - cartoon
  - anime
  - semi-realistic
  - text-to-image

Aurora - By Consistent Factor

Version 3, Version 3 Lumina, and Version 3 Chroma is here!

What is Aurora?

Aurora was made to counterbalance the realism that Euclid provided, giving you a model that can go from creating cartoons to anime and even creating semi-realistic images. As you can see from the preview image it does have a lot more freedom and artistic expression than Euclid. So please enjoy and have fun.

What is Aurora V3?

V3 is an updated version of Aurora which adds some more freedom to what V2 offered. I added a anime style model, a cartoon style model and added Dreamshaper 8 to the mix to compliment Dreamshaper 6 which was included in V2. V1 was a side project which was only really made for cartoon/anime type images and to see what I could do outside of the realms of realism. V2 expanded on that as V1 was quite limited and to be honest, rubbish. V2 was now able to cover pixel art and semi-realistic images as well as pretty much everything between to an extent. V3 expands on this further and also provides a better overall image, but does sort of strays away from overall cartoon/anime unless prompted to do so.

What is Aurora Lumina?

V3 Lumina is a more colourful and bold version of the base V3 model. It was designed to make everything pop and add more to the overall image, I also incorporated what I learned from Euclid Cinematic and applied that to this model to really let it loose.

What is Aurora Chroma?

V3 Chroma does the complete opposite of Lumina, instead of bright, bold colours with more going on, Chroma has less vivid colours and a more drawn/painted aesthetic. It would be perfect for those who like to create D&D characters to print out.

Style tags: verism style, samdoesart style, thepit style, owler style, cherrmous style, arosen style, uodenim style, stanleylau style

Prompt: I would use things like 16k, sharp focus, masterpiece, breathtaking, atmospheric, intricate natural lighting but it all depends on what style you are going for. Adding certain styles or adding things like cartoon, anime, 2.5D etc can all affect how the model creates the image, it can even do pixel art.

Negative prompt: (bad-hands-5:1.05), easynegative, [( NG_DeepNegative_V1_64T :0.9) :0.1], ng_deepnegative_v1_75t, worst quality, low quality, normal quality, child, blurry, deformed, disfigured, morbid, mutated, bad anatomy, bad art, (bad teeth, weird teeth, broken teeth), (worst quality, low quality, logo, text, watermark, username), incomplete, bad_prompt_version2

Recommend: DPM++ 2M Karras/DPM++ 2M SDE Karras/Euler, Steps: 15+, Hires fix 0.25 - 0.4, CFG 3 - 9.