Hi all,
I am not quite sure where the problem is but I suspect that John is referring to the potential for aliasing when downsampling the response to a low TR.
If that's the case, then John is correct, the downsampled time-series created by waver will have aliasing. However, you don't want to remove the aliasing because you are using these time series to model a signal that would be aliased too.
I hope that helps.
cheers,
z