Loading
Tem um workflow do ComfyUI? Gere uma documentação legível direto no navegador.Torne um workflow legível.Experimentar agora ↗
:heart: This model page as I'll be adding more tools over the next few days.
Fonte do modelo
Descrição da fonte
:heart: This model page as I'll be adding more tools over the next few days.
A collection of tools I've used in the past, and making available. Help to get great images that aren't dull and plain like every other portrait you see. These are embeddings so they don't use up extra vram, they don't change the style of your preferred model or LoRA and you can use as many as you like without muddying up your image .
Different checkpoint models will require different strengths to work, typically 0.2 to 1.8:
TOOLS:
Style-Action (coming shortly) A single embedding that creates a robust action scene with flow, movement, and great composition and lighting.
Style-GHL Golden Hour Lighting : Gives your image that low angle bright warm lighting. Often called the "golden hour" of a sunrise or sunset that gives the best artistic lighting.
Fontes
1 fonteArtefatos de modelo
1 artefatoTrechos de fonte
2 trechosSource context: 23076 downloads · Type TextualInversion · Base model SD 1.5
Style-DoF Depth of Field : Significantly increases the blur and bokeh of the foreground and background of your image. Helps your subject stand out. High settings can cause a image to look more photorealistic. Even higher will make it look like macro photography of miniatures.
Style-BlurFore Blurred Foreground : Great way to blur your foreground objects without increasing the overall DoF of the image, perfect for action scenes. You should prompt your desired foreground objects.
Style-Volumetric Volumetric Lighting : Adds a very cinematic haze, fog, particle effect to images that help give depth to distant objects.
Style-DA Dutch Angle : Camera tilt that gives the image an artistic, cinematic feel. Helps improve the sense of action and movement in your image. Works best with an action scene, prompt to help. (Warning: some models have difficulty distinguishing between dutch angle and dutch nationalities. While I've made every effort to remove those tokens from the embedding, you may still need to put "orange, dutch, holland or windmills" in your negative prompt. Also, sometimes SD will tilt the subject instead of the camera view, you'll need to cherry pick for best results, or use controlnets).
Style-Crp Cropped: Tight crop on images helps to create an artistic and cinematic composition when combined with other camera angles and action.
Style-EMCG Epic Movie Color Grading: A color grading that give a style similar to epic blockbuster movies, referred to as teal & orange. This embedding doesn't need much strength to work well. In fact, you will want to be sure and tag the colors you want so that it doesn't end up teal and orange if you don't want that. Adding blue, and teal to the negative prompts will help as well.
more camera angles to be added beyond this set
Do you have requests? I've been putting in many more hours lately with this. That's my problem, not yours. But if you'd like to tip me, buy me a beer. Beer encourages me to ignore work and make AI models instead. Tip and make a request. I'll give it a shot if I can. Here at Ko-Fi
Source context: 22983 downloads · Type TextualInversion · Base model SD 1.5
