SD3 Animation V1.1
5.0
0 reviewsDescription
Basically it is a simple deforum(with readable text) realistion in comfyui, I meant to maximise the functionality of SD3 Medium.
Input origenal image, run the workflow, the next images will be generated based on the previous one. After sevel generations, all the image will be warped up to form a video.
Change X&Y offset to control the camare direction.
Change image crop to determain zoom in scale.
Change denoise to modify the video consistency.
if you want "prompt travel", you need to manually change the clip text encoder, this workflow is not perfect.
Have fun building more on top of it.
-------------------------------------------------------------------------------------------------------------
在 comfyui 中实现的一个简单的 deforum(生成的视频中带可读文本),目的是最大限度地发挥 SD3 Medium 的功能。
输入原始图像,运行工作流,下一帧会根据上一帧生成。重复多次,所有图像将组合成一个视频。
改变 X 和 Y 偏移量来控制摄像机的拍摄方向。
更改图像裁剪以确定缩放比例。
改变去噪效果以修改视频的一致性。
如果你想要变换主体,你需要手动更改提示词,这个工作流程并不完美。
请尽情在此基础上开发更多功能。
Discussion
(No comments yet)
Loading...
Resources (1)
Reviews
No reviews yet
Versions (2)
- latest (a year ago)
- v20240617-073445
Node Details
Primitive Nodes (68)
Anything Everywhere (2)
GetNode (42)
ModelSamplingSD3 (1)
Prompts Everywhere (1)
SetNode (22)
Custom Nodes (113)
ComfyUI
- VAEDecode (21)
- VAEEncode (21)
- CheckpointLoaderSimple (1)
- PreviewImage (2)
- LoadImage (1)
- CLIPTextEncode (2)
- KSampler (21)
- ImageCrop+ (21)
- RIFE VFI (1)
- VHS_VideoCombine (1)
- ImageResizeKJ (20)
- ImageBatchMulti (1)
Model Details
Checkpoints (1)
sd3_medium_incl_clips_t5xxlfp8.safetensors
LoRAs (0)