The 96K filter design pretty much solve all the phase and overshoot anomaly! it is at 0 now. pretty accurate and smooth!!! I am happy!
But i have a new problem. The time windowing in the end introduce a dip in the final result.
To solve it, i need to increase filter delay! BUT when I do, the XO slope goes up, then i need to increase filter length that create the middle dip again!!! So to get balance between them is crucial.
Do you have tip how to make the magnitude flat, as some windowing will splice/cut it!!
I attach my measurement data, and setting, if you can, it will be great to see how you do it to create flat phase from 200hz with nice magnitude.