Download/Rating
Woafu's comment
Putting (wind:1.5) in the prompt basically tells AnimateDiff to focus on the background. I use it a loooooot as I don't want the script to try (and badly fail) too complex animations.
Image Generation
| UI | Stable Diffusion A1111 (External link to GitHub) |
|---|---|
| Checkpoint | RealCartoon-Anime |
| Character LORA | Chiori (External link to CivitAi) |
| Prompt | <lora:chiori:1> chiori, 1girl, brown hair, drill hair, hair ornament,hair flower, flower, red eyes, elbow gloves,short sleeves,wide sleeves, japanese clothes, kimono, pleated skirt,dress, black thighhighs, enigmatic smile,proud, confident,((Facing viewer)),(upper body,hands out of focus), (wind:1.5), three quarters view from above, sunrise,sakura blossoms,sakura trees,medieval japanese temple, 9:huge seductive smile, looking at viewer,blush |
| Negative prompt | (deviant art, deviantart,anime), (worst quality, low quality, ugly:1.4), poorly drawn face, out of frame, mutation, mutated, extra limbs, extra legs, extra arms, disfigured, deformed, cross-eye, blurry, (bad art, bad anatomy:1.4), blurred |
| Generation size | 432x768 |
| Steps | 60 |
| Sampler | DPM++ 2M SDE Karras |
| CFG Scale | 14 |
| Hires size | 520x920 |
| Hires steps | 30 |
| Hires upscaler | RealESRGAN_x4plus_anime_6B |
| Extensions | AnimateDiff |
| AnimateDiff Parameters | model: v3_sd15_mm.ckpt video_length: 16 fps: 24 loop_number: 0 closed_loop: R-P batch_size: 12 stride: 1 overlap: 3 interp: FILM interp_x: 3 |