Hello all, my name is Ian and I’m interested in the Compositor UI Improvements idea for GSoC 2025. I’m new to open source, but I’ve been using Blender for many years now and would love to contribute to this community. @izo and @OmarEmaraDev are the possible mentors for this project.
I’ve been looking over the documentation and relevant files for compositor nodes, so I have a basic grasp of how things are structured and how the current Crop and Corner Pin gizmos are functioning. I’ve come up with some ideas for further implementation and would love some feedback, considerations, or pointers of things to check out and study more.
Brief Synopsis
Currently, most compositor effects rely on manual node inputs, which makes repeated adjustments cumbersome. This project will expand the functionality of compositor nodes by implementing 2D gizmos and UI tweaks for transform nodes. By implementing interactive controls, users will be able to adjust node transformations directly from the compositor Backdrop. The goal is to have a straightforward, simple, and useful feature implemented by the end of the program.
Deliverables
Scale, Rotate, Translate, and Transform 2D gizmos
Show all Scale gizmo controls on hover
End user documentation
Potential additional tasks:
Blur/Effect region indicators or bounding boxes to visualize radius of effects
Flip gizmo
Clamp Value for Hue/Saturation/Value Node (Suggested by user EliotMack)
More Detailed Proposal:
Here is my WIP proposal. Please feel free to comment or add suggestions for improvements!
Please also feel free to reply with any critiques or suggestions!
I think the ideas are in principle good and align well with our priorities in the compositor. My concern is that the scope might be too large for a GSoC project.
I have the following comments:
Add a proposed size for the project: 90, 175 or 300 hours
You probably mean to add these gizmos in the compositor Backdrop, not the 3D viewport or the image editor. So it’s good to be specific.
There is already a patch for box mask gizmo, so it shouldn’t be proposed here.
“Always see” sounds nice, but would require significant changes to the current design. The proposal should contain more details on how that should work, at least from the user’s perspective. I would probably count this one as a GSoC project by itself.
I’ll definitely do that - the original idea listed either 90 or 175 hours depending on the scope. Cutting out the box mask/mask shape/“always see” gizmos, do you think the scale would be more manageable? I thought that it would be too little which is why I brainstormed extra haha.
Yes, great catch. I’ll definitely specify that.
Since it seems that I am overshooting the workload a bit, maybe trimming down the deliverables to just the transform gizmos and blur/effect region indicators would be better. I’ll boil down the extra ideas to 1 or 2 things as well.
The suggestion was a rough estimation for about 1-2 gizmo types only. Suggesting a 300 hours project is totally fine, as long as the estimation for deliverables is reasonable.