I plan to use Giskards modified 3145 passive network as a starting point for this project, and also use a DSP or digital EQ for room/speaker correction. Is it possible to use only the frequency dividing components in the network, leave the L-Pads out of the circuit, and use the DSP for levelling of the drivers, or is this just BS? Is it better to build complete networks and then add the DSP? Has anyone here tried this, or something similar?