Subdividing faces II

Hi all
referring to this old thread
I wanted to ask is anyone found the way to subdivide custom mesh inside vvvv?
i see the practical usability in mapping low-resolution mesh > post subdividing > option for finer deforming of geometry (attractor etc.)


Absolutely a job for DX11. Check examples in packs/dx11/girlpower/sm5/tesselation

Thanks. i forgot to say i need it for DX9


I’m also in the need for a DX9 version of subdivide faces. … Did you find a solution for that?

cheers, m9d

Why u would use dx9?

Still in love for DX9 … unfortunately I hadn’t the time yet to look into DX11 ;(

Well… u shud… Tesel is also in examples already. Don’t think it somehow possible in dx9 since u need to change topology bla-bla bla.

If you still want to do your rendering in dx9 you can build only the mesh in dx11 with the tesselation geometry shader and convert it again to a xmesh using vertexbuffer join etc

which node to use?

Yea it’s gonna be pretty hard, but it shud be readback, also u gonna need to redo ur indices

@ antokhio: indeed
so any hints what’s next?

I think colorsound made a module called geometry split or something like that- it does what you want iirc

but really if you want to play more with tessellation just switch to dx11. Going this other way with readback you are copying everything gpu>cpu>gpu, and losing the advantage of having hardware specificly made for doing exactly what you want at interactive rates.

yea the fps gonna be rather slow