Hi there,

Stupid question on spatial smoothing: does SPM compute a sort of mask 
for the brain before smoothing or simply smooth all data? (thus mixing 
up non brain and brain tissues on the edges)

Another, maybe less stupid ( :-\ ), question is: if I observe an effect 
in a small region (a nucleus) in VBM with a FWHM of 8mm but not at 4mm 
nor at 12mm ; Am I right to assume that's because my filter matches the 
signal? (at 12mm I may mix the data with non brain hence a drop of  the 
effect but at 4mm this should not - the observed effect is thus not 
related to smoothing mixing up data but related to the spatial size of 
the effect)

Any help will be greatly appreciated.

The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.