Quick OpenPose Character Concept Sheet Creator V1.2
5.0
0 reviewsDescription
What this workflow does
This is just a Quick Character Concept Sheet designer. It's Very simple, and I use it all the time - So I thought I'd Just share it with everyone.
It will create two OpenPose images based on your choice of head and body positions. One for SDXL Base OpenPose ControlNet model compatibility and one for SD base OpenPose ControlNet model compatibility. I personally find these sizes to be optimal.
✨ BONUS CONTENT ✨
I have added some extra workflows that can be found in the Resources section over on the right, that will give you an example of how to Utilise This workflow.
- Add / Use Your Own Image:
- This is an example Of how you can bring in your own characters
(more to come)
How to use this workflow
🟪 Face and Character Positions 🟪
Defaults are good, but you have some freedom here to pick the rotation and position of the head and body.
I added Just head and Just Body - Which will better aid if you want to bring in your own Character and use this to create their Character Concept Sheets. Basically you would IPAdapter the head and then the body separately - I find if you try to do these together then you get weird results in the head images. I will add examples to bonus content later.
🟩 Example Usage 🟩
Here is just a quick set of nodes thrown together as an example of what you can do.
🟫 Pipe In Examples 🟫
Just added a simple Pipe in Example of how the images and masks can be passed though and used.
Tips about this workflow
The fundamental part of this workflow is the 🟨 Section.
You can remove the 🟩 & 🟫 sections if you want too - They're just there as examples.
I have included output nodes for both images so you can copy this group into your own workflows (or even make it a template 😉) and then attach the desired output nodes to your OpenPose ControlNet Model workflow. The ideal size for the Base Latent image is in the title of each Image node:
- 2048 x 1024
- 1536 x 768
But of course, even though these are optimal sizes, you can resize them as desired.
Version Changes
v1.2
- Added Separate Head and Body Pose Creators
- Re-worked Bonus Content - Add Own Image
v1.1
- No Major updates to main Group Node
- Amended some stuff in Example Usage Group
- Added Pipe In Example Group
- Added Bonus Example "Add / Use Your Own Image" (as requested)
v1.0
- Release Workflow
Discussion
(No comments yet)
Loading...
Resources (1)
Reviews
No reviews yet
Versions (3)
- latest (2 months ago)
- v20241014-185735
- v20240920-205804
Node Details
Primitive Nodes (20)
DownloadAndLoadLivePortraitModels
ImageOverlayResized
LayerMask: LoadFlorence2Model
LayerUtility: Florence2Image2Prompt
LivePortraitCropper
LivePortraitLoadCropper
Note
Reroute
Switch conditioning [Crystools]
workflow>Model / Prompt Passthought
workflow>Model/Clip Passthrough
workflow>Seed + Sampler
Custom Nodes (119)
- ImageBatchJoin
- ImageTransformPaddingRelative
- ImageTransformResizeRelative
- CR Color Panel
ComfyUI
- PreviewImage
- ImageToMask
- ImageBatch
- LoadImage
- CLIPTextEncode
- EmptyLatentImage
- VAEDecode
- ControlNetApplyAdvanced
- ConditioningZeroOut
- KSamplerAdvanced
- CheckpointLoaderSimple
- ControlNetLoader
- easy joinImageBatch
- MaskBoundingBox+
- GetImageSize+
- ToBasicPipe
- LayerUtility: ImageScaleByAspectRatio
- LayerMask: SegmentAnythingUltra V2
- LayerUtility: ImageRemoveAlpha
- LayerMask: SegmentAnythingUltra
- IPAdapterFaceID
- IPAdapterUnifiedLoader
- IPAdapterUnifiedLoaderFaceID
- IPAdapterNoise
- IPAdapterTiled
- Cozy Pose Face Reference
- Cozy Pose Body Reference
- JoinStrings
- ImageGridComposite2x2
- ImageResizeKJ
- GetImageSizeAndCount
- ShowText|pysssss
- MathExpression|pysssss
- Text _O
- Image Remove Background (Alpha)
- Text to Conditioning
Model Details
Checkpoints (1)
SD - MERGE\L\lifeLikeDiffusionEthnicitiesSupportedNative_lifeLikeDiffusionV30.safetensors
LoRAs (0)