Automating quad topology for dense meshes (workflow discussion)

anvilinteractive

I’ve been working on quad remeshing workflows for animation and game assets, specifically around edge flow consistency and deformation safety.

Most automatic remeshers are fine for previews, but I kept running into problems on denser meshes where diagonals and uneven flow made rigging unpredictable.

Below is a before/after example showing the kind of topology issues I was trying to solve.

The focus here is not speed, but producing quads that hold up under animation and deformation.

I ended up building a Blender add-on to support this workflow, but I’m mainly posting to get feedback from others dealing with similar topology problems.

Curious how others here approach quad remeshing for animation-heavy assets, especially on dense or organic meshes.

If anyone wants to see the exact setup or workflow I’m using, I documented it here:

https://github.com/willyanibe/quadify-blender-remesher

Eric Chadwick

Can you show some more examples? Maybe this car: https://github.com/KhronosGroup/glTF-Sample-Assets/tree/main/Models/CarConcept#car-concept

Also it's a bit hard to see the topology, could you use shaded wireframe instead? Like this: 

sacboi

On the face of it an interesting closed source option however I think a brief workflow overview would help, thus enabling people to assess either it's pros or cons themselves although appears your demo is no longer available?

Side note - for hard surface stuff I make, a native voxel remesh technique seems straightforward (tho in that vid rather than HardOps I'd use WNs) enough without having to implement 3rd party alternatives:

https://www.youtube.com/watch?v=hC7EeqhKiak

Eric Chadwick

Also, the images on GitHub were created with GenAI and show several significant errors. Like spelling, and ignorance of actual topologies. They don’t show what the product actually does! This makes any user who knows topology not want to buy this product.

The 2nd YouTube link in GitHub is broken, as stated. But also, digging through your YouTube videos, there isn’t a clear workflow video that explains how the buttons all work.

It would help to have actual documentation explaining what each button does.

anvilinteractive

Thanks for the feedback, guys. You can test out the base version here https://anvilinteractive.gumroad.com/l/quadify-smart-quad-remesher-for-blender
Please feel free to test out different models on the base variant (it is free), and upload your videos!
I will provide several other videos of different test scenarios for it.
Also, @Eric Chadwick, I will fix the issues you outlined

anvilinteractive

sacboi said:

On the face of it an interesting closed source option however I think a brief workflow overview would help, thus enabling people to assess either it's pros or cons themselves although appears your demo is no longer available?

Side note - for hard surface stuff I make, a native voxel remesh technique seems straightforward (tho in that vid rather than HardOps I'd use WNs) enough without having to implement 3rd party alternatives:

https://www.youtube.com/watch?v=hC7EeqhKiak

I quadified your car model and created a youtube video of it.
Check here: https://youtu.be/FZAtqGWD7-g

sacboi

Not that I'm wholly adverse too individuals or company entities promoting pipeline tools, which might be viable to integrate into ones workflow but further research indicates Anvil Interactive Solutions leverages AI support developing their product. Now as for myself again neither here nor there what others choose to work with generating content, however for future reference. Instead of omitting certain info, I do recommend full disclosure upfront hence allows the community to make a holistic determination, that may or may not align with personal choice.     

anvilinteractive

sacboi said:

Not that I'm wholly adverse too individuals or company entities promoting pipeline tools, which might be viable to integrate into ones workflow but further research indicates Anvil Interactive Solutions leverages AI support developing their product. Now as for myself again neither here nor there what others choose to work with generating content, however for future reference. Instead of omitting certain info, I do recommend full disclosure upfront hence allows the community to make a holistic determination, that may or may not align with personal choice.     

You are right on both counts.
Thanks for pointing that out, it will be made clear here on out.