dose calculation

Hi,
I am replying to this post as I am trying understand the way to get the actual dose rate in [pSv/s] from tallies. The normalization with respect to volume or source seems to be trivial.
From this post it seems that using EnergyFunctionFilter with dose_coefficients would modify the unit of the tally?
Otherwise, if the dose is obtained in post-processing using (flux tally)*(dose-to-flux-coefficient tally), how can I quantify the uncertainty on the dose? I guess there can be a strong correlation between the tallies if I use the same volume/mesh, so I can’t simply sum up the variances