Volume plots with vtr file

[ikaul00]

Hi everyone,

I have to make a visualization of a 3d volume rendering of density data on a cartesian grid. My data is initially in the form of a 3d numpy array but in spherical coordinates (with a statically refined mesh). I set out to first convert the numpy array and the dimensions to a .vtr file using pyevtk with the function gridToVTK. I then load it into the Visit GUI and apply the spherical to cartesian transform. However the full plot looks like a slightly transparent sphere, which seems off because the volume rendering should not look like that (the density is 4 orders of magnitude higher in the midplane than the outermost regions). More importantly, the rendering looks like it is only giving specific values at the points I have stores the data for, instead of making smoother surfaces to 'connect' these point data. I tried different combinations of the opacity settings as well but the 'pointiness' does not go away. Is there something wrong with the way I store the data and load it into Visit?

Any help would be appreciated!
Thanks

[biagas]

Could you provide the .vtr file (zip and attach here if you are able/willing).
Also, the exact options you used for the Transform operator and Volume plot.
Which version of VisIt are you using and which OS are you running?

[ikaul00]

Thank you for the reply!
The zipped file is ~100MB and so github does not allow me to upload it (anything >25MB) - is there any other way to upload it?. But I am using Visit 3.3.0 on MacOS Big Sur 11.4. I went to Operator Attributes -> transform -> coordinates and chose spherical (data) to cartesian (to visualize). The volume plot has Default rendering, and 50000000 sample points, with default lighting. Color is on a log scale Min 1e-15, Max 7e-12, with opacity starting at 0 (at the beginning of the color spectrum) and rising up uniformly to the maximum at the end of the color spectrum, with the default opacity variable.

[markcmiller86]

is there any other way to upload it?.

Have a look at anon-ftp method here