# about contour filter
**URL:** https://discourse.paraview.org/t/about-contour-filter/14043
**Category:** ParaView Support
**Tags:** vtk
**Created:** [February 28, 2024, 9:58am UTC](https://discourse.paraview.org/t/about-contour-filter/14043 "2024-02-28T09:58:42Z")
**Posts on this page:** 7
**Page:** 1
### Author:  [@865979216](https://discourse.paraview.org/u/865979216)
#### Post date: [February 28, 2024, 9:58am UTC](https://discourse.paraview.org/t/about-contour-filter/14043/1 "2024-02-28T09:58:42Z")
i have a .mha file，then i use vtk.vtkImageMarchingCubes() to get marchingCubes.SetValue(0, 39)，the result is bad
but i use Paraview Contour
the result is better than vtk.vtkImageMarchingCubes()。
why？
---
### Author:  [@cory.quammen](https://discourse.paraview.org/u/cory.quammen)
#### Post date: [February 28, 2024, 4:31pm UTC](https://discourse.paraview.org/t/about-contour-filter/14043/2 "2024-02-28T16:31:30Z")
ParaView uses `vtk.vtkContourFilter` which dispatches to a number of different contour filters that perform the actual contouring. To my knowledge, it does not dispatch to `vtk.vtkImageMarchingCubes`.
Try using `vtk.vtkContourFilter` in VTK and see how the results match. You should be able to simply replace `vtkImageMarchingCubes`.
---
### Author:  [@865979216](https://discourse.paraview.org/u/865979216)
#### Post date: [February 29, 2024, 8:06am UTC](https://discourse.paraview.org/t/about-contour-filter/14043/3 "2024-02-29T08:06:18Z")
the code is
the result is also bad ，just same with vtk.vtkImageMarchingCubes，
。
But ，I find “PVContourFilter”
Maybe，“PVContourFilter” is the key 。I want to get the result like Paraview’s Contour，but I don’t know how to code。 [ParaView/VTKExtensions/FiltersGeneral/vtkPVContourFilter.h at master · Kitware/ParaView (github.com)](https://github.com/Kitware/ParaView/blob/master/VTKExtensions/FiltersGeneral/vtkPVContourFilter.h)
---
### Author:  [@865979216](https://discourse.paraview.org/u/865979216)
#### Post date: [March 1, 2024, 10:34am UTC](https://discourse.paraview.org/t/about-contour-filter/14043/4 "2024-03-01T10:34:36Z")
Do u have some ideas？
---
### Author:  [@cory.quammen](https://discourse.paraview.org/u/cory.quammen)
#### Post date: [March 1, 2024, 2:05pm UTC](https://discourse.paraview.org/t/about-contour-filter/14043/5 "2024-03-01T14:05:42Z")
`vtk.vtkPVContourFilter` is in the ParaView source code, so it isn’t available with just VTK. You could use ParaView’s `pvpython` application and do your processing there if all you care about is the output isosurface, but I suspect you care about more.
As to why the same isovalue produces different results with the different filters, I’m at a bit of a loss. Is it possible to share your data?
---
### Author:  [@865979216](https://discourse.paraview.org/u/865979216)
#### Post date: [March 4, 2024, 2:38am UTC](https://discourse.paraview.org/t/about-contour-filter/14043/6 "2024-03-04T02:38:32Z")
[VascularEnhance.rar](https://discourse.paraview.org/uploads/short-url/e7MbPu5XohZTZPDMO47EjMDX1xF.rar) (808.9 KB)
This is the MRA data。So far, I only care about outputting results because my supervisor asked me to do research on vascular analysis. I use mha（dicom is also ok） to output vascular vtp files, and then use vmtk to generate vascular centerline vtp files。So，I want vascular vtp files to look good
---
### Author:  [@865979216](https://discourse.paraview.org/u/865979216)
#### Post date: [March 4, 2024, 12:03pm UTC](https://discourse.paraview.org/t/about-contour-filter/14043/7 "2024-03-04T12:03:33Z")
sorry，I know the reason。I downsampled the data 😵💫