Hi Romi,
It seems to be a bug in FDATool. The display panel is giving out the wrong answer. The designed filter is indeed a filter of order 20. If you look at the coefficients or the impulse response, there are indeed 21 coefficients. Thanks for reporting the issue.

