BTW, I am not suggesting that this research is a long term solution for deleting Cinavia, but it is necessary investigation into understanding how Cinavia works in the real world. The first step is to figure out how to delete it once.
You can't see the low frequency modulation difference in the graphs. The BD audio definitely has a larger higher average amplitude and wider bandwidth. The only way I noticed it was by listening in the headphones.
My next steps are to normalize the relative amplitudes. Then apply a constant factor to the BD audio to equalize it with the DVD audio's amplitude. Then invert the BD audio and sum it with the DVD audio to derive the Cinavia signal itself.
I can't think of any theoretical reason why this low frequency amplitude difference would occur with different mastering techniques, but the reason for it is trivial once you know it's there, you can compensate for it.