# Contour filter for high-order elements

**URL:** https://discourse.paraview.org/t/contour-filter-for-high-order-elements/7001
**Category:** ParaView Support
**Tags:** high-order-elements
**Created:** [April 19, 2021, 4:22pm UTC](https://discourse.paraview.org/t/contour-filter-for-high-order-elements/7001 "2021-04-19T16:22:49Z")
**Posts on this page:** 1
**Page:** 1

<div class="post-metadata">

### Author: ![nvmr](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/n/df705f/32.png) [@nvmr](https://discourse.paraview.org/u/nvmr)
#### Post date: [April 19, 2021, 4:22pm UTC](https://discourse.paraview.org/t/contour-filter-for-high-order-elements/7001/1 "2021-04-19T16:22:49Z")

</div>

Dear ParaView-Support,

I try to implement the vtkLagrange\*Cells into a high-order DG code in order to visualize the solution without requiring an extra internal interpolation procedure. After some problems in the beginning, it seems that I made it work. At least the subdivision refinement works as expected.

However, if I create a contour from the high-order representation, the contour seems to be linear. ![grafik](https://discourse.paraview.org/uploads/default/original/2X/e/e0f921f5268eb5f17f5f85e85c4cc529d5ae9262.jpeg) .

Increasing the number of subdivisions has no effect. Using the tessellation filter improves the representation, but it takes a significant amount of time to apply the filter (with \>4 subdivisions).

 ![grafik](https://discourse.paraview.org/uploads/default/original/2X/2/2036be782085b011b40a981992e28bab16a37875.jpeg)  
(nSubdivisions=4, not optimal as well as the solution is continuous)

So, is there an alternative way to increase the resolution of high-order elements in a practical amount of time? Do I miss some option/filter to do that?

Thanks!  
Michael