I've never been able to wrap my mind around solving the faces individually. After all, since there's no standard color scheme I'm basically doing that when I solve it like a 4x4x4.
So you're saying that rather than one center being wrong, there are 4 others that are actually wrong? How do I know which four?
Maybe, a picture would help that shows the exact situation?
You wrote earlier:
I've tried these techniques to solve in the past and it's worked. However, I've come across a new case. Everything is solved, except ONE center is rotated 90 degrees. All fisher cube and 4x4x4 supercube algorithms I've used have not worked, because they only work when two are rotated 90 degrees.
- ONE center means a centre of the 4x4x4 consisting of FOUR centre pieces (tips of the Octahedron). (I do not see a different interpretation of your sentence.)
- If you turn those 4 centre tips( 4 centre pieces building the centre of a face on the 4x4) of the Octahedron by 90 degrees all 4 centre are solved, right?
If my understanding was correct, just do that and afterwards put the edges (trapezoids) and corners (little triangles) of the 4x4 back to their destination.
Obviously, there are other solutions as well, but this is the easiest to describe.IF
i have understood your description correctly, that is!
Still, I believe strongly that solving it face by face is much easier. You just need two commutators (see this other thread), easy to understand and easy to be memorized.
Afterwards, you handle it as a 2x2 and can produce any colour scheme you like.