You can simply type cast each value to I16 (in the Advanced -> Data Manipulation function palette). Maybe there is a better way, it's hard to say without taking a look at the code.
However, if I can judge upon your graphs, there's another problem: it seems that your output data span a range greater than 64K, while input is limited to about (-4000, 4000) - why ? is there any kind of rescaling ?
Those output data cannot be represented with I16, you should use I32 instead (not U32 as erroneously written in my first reply, as you need negative values).
Paolo
Paolo
-------------------
LV 7.1, 2011, 2017, 2019, 2021