Hi Frank,
Your results seems to be very convincing. I sure would like to see my mount, CGEM-DX, to be able to converge. It seems to get stuck in the 80 pixel area. I hope that sometime in the near future Jared or Ken can look into this further.
In the meantime, I'm not a computer expert, is there something you might be able to help us folks out that would allow us to execute your code that would allow for more precise centering when centering is called for in SGP? Or was your code simply to show how this new center algorithm might be able to be implemented?
Thanks,
mahaffm