Spatial filtering of the field data during postprocessing

I have a set of unsteady LES data, and I would like to filter out (low-pass) the higher frequencies over all timesteps. Essentially, what that would do is smooth out the disturbances in a scalar or vector field over all instances of time. I am wondering how this operation cab be performed in ParaView (5.9.1). I can’t seem to find a lowpass or a bandpass filter option. Would I just have to rerun the case at lower mesh resolutions?

For some clarity on the above query, below is some example representation of the unsteady velocity field filtered at 50, 10 and 1kHz from top to bottom (with the bottom most figure showing smoothed out contours).

Does the Smooth filters works for you ?

Not really sure what the smooth filter is doing. It seems running a smaller number of smoothing passes does not show me any significant change in data, and running higher number of passes distorts the data, like the comparison below (original data and smoothed data). Also, I can’t seem to know what frequencies I am cutting off with this filter.


Original data

Smoothed

Indeed, this doesnt look like the expected result.

I’m afraid spatial frequency filtering is not a thing in VTK. You may be able to do some similar thing using a gradient filter and some targeted smoothing but this is definitely not easy.