Though Elite framework not supporting the floating point processing module in an audio topology, user can add floating point processing modules in audio topology.
Floating point processing module (PP Module 2) can accept the data that is coming from PP Module 1 in Q15 format, treat it as an equivalent floating point number and it can process the data or user can have the mechanism to convert the fixed point data to floating point data before its going to floating point processing module.
If there is a floating point processing module next to the current floating point processing module then it doesn't make sense to convert data back and forth between fixed and floating point between these two floating point modules. So these two modules can be wrapped under one wrapper and treat them as a single floating point module as shown in above figure.
Finally before the output of the floating point processing module going to next fixed point processing module, floating point data need to be converted back to fixed point (in this case it should be in Q15 format).
We recommend to have the conversion process from fixed to float and float to fixed point data with in floating point processing module before starting the data processing and after processing the data respectively