# Nonlinear subdivision for rational Bézier curves not working
**URL:** https://discourse.paraview.org/t/nonlinear-subdivision-for-rational-bezier-curves-not-working/6522
**Category:** ParaView Support
**Tags:** vtk
**Created:** [February 22, 2021, 1:33pm UTC](https://discourse.paraview.org/t/nonlinear-subdivision-for-rational-bezier-curves-not-working/6522 "2021-02-22T13:33:00Z")
**Posts on this page:** 3
**Page:** 1
### Author:  [@JHarsch](https://discourse.paraview.org/u/JHarsch)
#### Post date: [February 22, 2021, 1:33pm UTC](https://discourse.paraview.org/t/nonlinear-subdivision-for-rational-bezier-curves-not-working/6522/1 "2021-02-22T13:33:00Z")
As written in the [blog post](https://blog.kitware.com/implementation-of-rational-bezier-cells-into-vtk/) vtk recently supports rational Bézier cells. The two and three dimensional cells are visualized as expected by ParaView but the curve elements seem to be broken.
- Beziér curve with nonlinear subdivision level 4 is not rendered as expected. Only the control polygon is visible:
- Using as hotfix, the tesselate filter results in the expected visualization:
The VTK source file is [VTK\_BEZIER\_CURVE\_quadratic\_quarter\_circle.vtu](https://discourse.paraview.org/uploads/short-url/mmOskJCuEXU978L1GcgdCfSopxO.vtu).
Examplary, the surface element [VTK\_BEZIER\_QUADRILATERAL\_Piegl\_figure\_5.20.vtu](https://discourse.paraview.org/uploads/short-url/avcrFY01QDgyi45kVKflmfcvm1s.vtu) works as expected, even for 2D surfaces embedded in 3D space:
I’m using ParaView 5.9.0 on a Windows 10 machine.
---
### Author:  [@florian360](https://discourse.paraview.org/u/florian360)
#### Post date: [February 22, 2021, 9:47pm UTC](https://discourse.paraview.org/t/nonlinear-subdivision-for-rational-bezier-curves-not-working/6522/2 "2021-02-22T21:47:15Z")
Yes, this is a known issue on our side and you already found the workaround with the tessellation filter. However, this issue is not yet reported here [https://gitlab.kitware.com/vtk/vtk/-/issues](https://gitlab.kitware.com/vtk/vtk/-/issues) so feel free to do so.
You will note that you get the same problem with Lagrange curves. The problem is that in vtkDataSetSurfaceFilter, the nonlinear subdivisions are only implemented for surfaces, but not yet for curves.
---
### Author:  [@JHarsch](https://discourse.paraview.org/u/JHarsch)
#### Post date: [March 1, 2021, 6:39am UTC](https://discourse.paraview.org/t/nonlinear-subdivision-for-rational-bezier-curves-not-working/6522/3 "2021-03-01T06:39:18Z")
Thank you for the answer, I will write an vtk issue and hope for the best.