Oversampling a Channel Strip - questions about OS methods, placements and signal paths

Moonbot7000

Im building a channel strip and exploring oversampling for the EQ portion. Ive seen the best results so far using the Hermite 2x macros and the "cubic ILO" macros. They both seem to cut a certain amount of aliasing out in their own right (bear in mind im driving the hell out everything with FM and obscene amounts of gain and saturation for testing purposes) but I heard the best results when the two were combined using the below method. First, is this even the correct way to implement oversampling regarding the placement of the respective macros, secondly, is using the two combined methods just overkill, or is the improvement in quality I'm observing considered a desirable trade off for, honestly, not a huge increase in CPU consumption?

Lastly, as gain is increased in the module, I am pushing more of the signal into the saturation modules to keep levels reasonable and model natural saturation of a channel strip. My current configuration has saturation after each of the 3 EQs in the 3 band EQ. Attached is also an .Ens if anyone would be so kind to check it out and provide some general feedback. Thanks!

EQ oversampled.ens