HMSG-ComfyUI全景图工作流
created 6 months ago
upscale
lora
39 nodes

886

82

Outputs
example image output 0 for workflow.
example image output 1 for workflow.
example image output 2 for workflow.
example image output 3 for workflow.
example image output 4 for workflow.
example image output optimized_comfy_workflows_user_uploads/7c49c55e-187c-4ce2-8fdf-ed635981ed1e/assets/ComfyUI_00026_.webp for workflow.
Description

图片生成后会有轻微接缝,无法接受拼接缝,可以使用SDwebUI绘制全景图。

流程中使用两次 SD放大,第一次放大跑100轮,开启第二次放大需要跑300轮时间很久可输出4096*2048分辨率图片,最后的锐化可以自定义设置。

采样器CF值可适当调大可以增加清晰度,初始分辨率2048*1024清晰度比1024*512要好很多。

修正XY轴采样器Asymmetric_Tiling_KSampler的denoise竟可能小可以保证图片重绘度。

Built-in nodes
Custom nodes
CLIPTextEncode
VAEDecode
PreviewImage
VAEEncode
UpscaleModelLoader
UltimateSDUpscale
Upscale Model Loader
SaveImage
LoraLoader
Seed
CheckpointLoaderSimple
SDXLPromptStyler
KSamplerAdvanced
EmptyLatentImage
Asymmetric_Tiling_KSampler
ImageSharpen
Custom files
new

These are separate files that the creator has uploaded for this workflow.

custom file comfy_workflows_user_uploads/7c49c55e-187c-4ce2-8fdf-ed635981ed1e/custom_files/微信截图_20240609230348.png

微信截图_20240609230348.png

Models
0
0
0
0