Prompting Tools:
weloveai.ca/stable-diffusion-... -- SD Artist guide
promptomania.com/stable-diffu... -- SD Prompt Builder
www.the-ai-art.com/modifiers -- Modifiers guide
Image Generators:
getimg.ai/ref/j6220jy0G -- Get 100 credits extra with this link
civitai.com/login?ref_code=__...
civitai.com/login?ref_code=__... -- Use any of these two codes for extra credits
playground.com/create -- Useful for Experimenting with prompts and easily accesing new models.
app.leonardo.ai/ -- All round best SD Platform, we can train custom models, access community models, and has advanced inpainting and outpainting features.
aqualxx.github.io/stable-ui/ -- Best Platform for testing SD Models, has an extensive array of popular models, has advanced features like hiresfix, samplers, clipskip
sd.pencilcase.ai/ -- Has Automatic1111 UI, support for prompt weightage, blending faces.
stablediffusion.gigantic.work/ -- Has support for prompt weightage.
Tutorials and Resources:
stable-diffusion-art.com/prom... -- Extensive tutorial on prompting, prompt weightage, keyword blending, BREAK keyword, how to get consistent faces, very helpful.
/ stable_diffusion_keywo... -- List of Good Keywords to use.
A helpful KZitem channel:
@siliconthaumaturgy7593 -- SiliconThamaturgy
/ channel
His videos on SDXL models, Stable Diffusion samplers, and Best SD1.5 Model for hands were very helpful.
NOTE : Civit AI, Easy Diffusion, and Pencilcase.ai have Prompt weightage as well as prompt scheduling.
Best Keywords for Stable Diffusion:
Canon50 ; DSLR : Makes the image look photorealistic, high quality
Rendered by Octane : 3D animation style, amazing lighting, SDXL
4K : Increases detail and image quality
By Art Germ : Amazing keyword, artistic improvements, high contrast and vibrance
Portrait : Increases detail in face and skin, produces close-up
Hyper-realistic : Increases detail, Improves clothing
Dramatic : Improves lighting, and face expression
Rim lighting : Improves lighting and contrast; specifying color is better
Crisp image : Makes the image have more detail and less blurry
Name of Artist
Best Models for Playground AI (And most other platforms):
Realism: NightVision XL (SDXL), Realistic Vision (SD1.5) - Good for portraits, objects, and landscapes
Digital art: Dreamshaper XL (SDXL), Dreamshaper 8 (SD1.5) - Dreamshaper is good for close-ups of subjects, especially humans but also animals and other objects.
Fantasy: Mysterious V4 (SDXL), ReVAnimated (SD1.5) - Good for all round fantasy, Mysterious V4 is especially good for a blend between anime and fantasy.
Anime: Counterfeit XL V1 (SDXL), Counterfeit V3 (SD1.5)
NOTE: Dynavision XL, and Deliberate 1.1 (not 2) are also good models but they are not in Playground, if you are comfortable using Civit AI, I reccomend checking out Dynavision XL, which is especially good for generating stylized or 3D renders.
Model reccomendations (Leonardo AI)
Realism: Leonardo Signature (SD1.5), Absolute Reality (SD1.5), Leonardo Vision XL
Artistic: Deliberate 1.1 (SD1.5), Leonardo Diffusion (SD2.1), Leonardo Diffusion XL, Leonardo Signature (SD1.5)
Digital Art: Dreamshaper 7 (SD1.5), AlbedoBase XL
Landscapes: Leonardo Signature (SD1.5), Dreamshaper 7 (SD1.5)
Portraits: Dreamshaper 7, AlbedoBase Xl, RPG V5, Leonardo Signature, Photoreal
Illustrations & Design: Leonardo Signature (SD1.5), Leonardo Diffusion (SD2.1)
3D Animation: 3D Animation Style
Anime: Anime Pastel Dream
Best Upscaler : krea.ai
Other Image tools : clipdrop.co
Негізгі бет Ғылым және технология Stable Diffusion Ultimate Guide. How to write better prompts, and use Image to Image, Control Net.
Пікірлер: 2