Scaling in the Mixed Radix FFT is controlled by the TP_SHIFT parameter which describes how many binary places by which to shift the result to the right, i.e., only power-of-2 scaling values are supported. Scaling is applied after the last rank has been computed. Internal calculations are unscaled, so a data type of cint32 is used internally to allow for bit-growth from an input of cint16. No larger type than cint32 is available, so if TT_DATA is cint32, you must avoid arithmetic overflow by ensuring that the data value range on the input is restricted to prevent overflow.