EthernalDope allows you to generate 3d or 2d images with a unique and charming design.
Also, if you are a ComfyUi user and have ever wanted to try generating videos with the MMD2Vid method, EthernalDope offers you very consistent and stable results
Although the model is designed to be used to generate 3D images, it can also generate nice 2d results. The versatility of the model allows you to obtain images with a recognizable style, although this will depend on how the LORA you want to use has been trained.
Recommended settings:
Sampler: DPM++ 2M Karras
Sampling Steps: 20 -25
Upscaler: 4x-AnimeSharp / R-ESRGAN 4x+ Anime6B
Hires steps: 10-18
Denoising Strength: 0,4-0,6
VAE: YOZORA
Recommended prompt order: (For better visualization, check the generation info in the images themselves)
Prompt:
(trigger word, lora)
BREAK (Description of the character's physical features)
BREAK (Description of the character´s clothes and accessories)
BREAK (Background description and poses)
BREAK (Image composition/camera angle)
BREAK (Final and specific details, extra LORAs or LYCORIS)
Negative prompt:
EasyNegative, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, (worst quality:1.2), low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry, lowres graffiti, (low quality lowres simple background:1.1), eyes
Recommended hands Lycoris for better results!
Have fun testing this model, any suggestion will be welcome. Please wait for a better version and more models.
