I thought of this combination, to try writing fsigs to a ft, then use that ft to hit the masses in a scanu, and resynthesize the result.  Pretty hard.
This .csd doesn't make any sound (input and output working fine), with this .matrix.  Tried a few settings, not sure if I'm doing what I think I'm doing though.
I didn't make any use of the k-flags of pvsftw, since I want the resynthesis to be continuous.
The only question I can think of is, will scanu continue to read from the "init" table for the shape of the hammer, or only at init-time?  The argument itself is i-rate, but I assume that just means one can't change ftables during performance.
Thanks.
-Chuckk

--
http://www.badmuthahubbard.com