Page 1 of 1

stellar drift alignment... (for when you can't see polaris)

Posted: Tue Oct 13, 2020 1:20 pm
by fjr1234
I'm sure I'm not alone in this; from my yard my view of Polaris is blocked. So I'd like to suggest adding a stellar drift alignment tool similar to the polar alignment one.

Thank's for your time.


Re: stellar drift alignment... (for when you can't see polaris)

Posted: Fri Oct 23, 2020 2:26 pm
by calypsob
Ive been curious about polar alignment accuracy in this regard. I wonder how far away from polaris you can polar align via plate solve before a noticeable drop in performance occurs.

Re: stellar drift alignment... (for when you can't see polaris)

Posted: Fri Oct 23, 2020 8:15 pm
by admin

I haven't actually sat down and done the mathematics to work out how the accuracy would drop off with distance, but I suspect that if both the position of the celestial pole and the current centre of rotation was significantly out of view of the camera then the results would start to degrade. It would be like trying to measure the distance between two plants in your garden by triangulating from the other end of the street – you'll get some sort of answer but I wouldn't want to bet on it being an accurate one.

In practice, you hit other limitations first – any more than about 6 or 7° out from the pole and you can no longer ignore spherical geometry which would significantly complicate the calculations that need to be performed. The number of stars you need to have in your database also multiplies by a factor of four every time you double the distance you want to cover the pole. SharpCap currently has about 15 to 16,000 stars per hemisphere to cover the area out to 7° from the pole – if you try to go to 15° out then you would need around 60,000 stars in the catalogue.

I've tried at various times to put together routines that don't require review of the celestial pole and have never come up with anything that I'm satisfied with. My criteria for a satisfactory polar alignment routine are (at minimum) that it should require go to movements as any inaccuracy in those will lead to inaccuracy in the final results and also you should be able to adjust both axes at once rather than having to adjust them separately.

Cheers, Robin