Hi,
When making a mask, the soft edge is applied such that the mask value is one at the initial binarization threshold, and decreases outwards to zero over the selected number of pixels. When making an inverse mask, it does the opposite - starting at 1 at the binarization threshold and decreasing inwards to zero. I would suggest it would be better if instead, when creating an inverse mask, the mask value started at zero at the binarization threshold, and increased outwards to 1 - otherwise it is hard to set what the appropriate threshold should be to mask out the desired volume (e.g. for the purposes of signal subtraction).
Cheers
Oli