# vtk file for a Face Cubic Centered distribution points

**URL:** https://discourse.paraview.org/t/vtk-file-for-a-face-cubic-centered-distribution-points/604
**Category:** ParaView Support
**Created:** [September 19, 2018, 7:30pm UTC](https://discourse.paraview.org/t/vtk-file-for-a-face-cubic-centered-distribution-points/604 "2018-09-19T19:30:57Z")
**Posts on this page:** 1
**Page:** 1

<div class="post-metadata">

### Author: ![Denisse\_Reyes](https://discourse.paraview.org/letter_avatar_proxy/v4/letter/d/e495f1/32.png) [@Denisse\_Reyes](https://discourse.paraview.org/u/Denisse_Reyes)
#### Post date: [September 19, 2018, 7:30pm UTC](https://discourse.paraview.org/t/vtk-file-for-a-face-cubic-centered-distribution-points/604/1 "2018-09-19T19:30:57Z")

</div>

Hello,

I’m using a Face Cubic Centered (FCC) distribution point instead of the usual Cubic distribution. But, when I plot an isosurface (Contour), it doesn’t look good.  
At first, for each (i,j,k) I’m using the following expressions for generating the mesh points:  
x1 = i; y1 = j; z1 = k  
x2 = i+0.5; y2 = j+0.5; z2 = k  
x3 = i; y3 = j+0.5; z3 = k+0.5  
x4 = i+0.5; y4 = j; z4 = k+0.5

So I have four points for each (i,j,k).

Secondly, I generated the vtk file using **_ni_** the number of cells associated to the index _ **i** _ and the same for _ **nj** _ and _ **nk** _ (with a STRUCTURED\_GRID):  
…  
`DATASET STRUCTURED_GRID`  
`DIMENSIONS ni,nj,nk`  
`POINTS 4*(ni)*(nj)*(nk) float`  
…  
The distribution point looks well when I use the paraview. However, as I mentioned before the isosurface doesn’t look good. I have been read about it and I found that is related with the number of cells `(ni*nj*nk)`, isn’t it?

Finally, I also tried to associate one (i,j,k) for each point in order to increase the number of cells. The contour looks better with this, but it doesn’t work at all.

So, do you know if there is some way to generate the vtk file? Do I need to use an UNSTRUCTURED\_GRID?

Many thanks,

Denisse R.