I am under the impression that I want the focus point for the flat to be the same as for the lights. I realize there is some tolerance, so the dust balls are the same siize as your lights. I use a flat panel directly on top of my dew shield (Gerd Neuman E-panel).
Over the course of a night my focus point may start at 280 and move to 232 over the course of 8 hours. I think this is pretty extreme and in some seasons the differences are even larger. I suspect this may have an impact on my flat quality.
So, in my example, if it moves by 50 steps over the 8 hours through 4 filters at ten twelve minute exposures each. (assuming purely linear change for my example)
- The first filter may be from 280 to 268 as the focus point
- The second from 268 to 256
- The third from 256 to 244
- And the last from 244 to 232.
The next morning when I do flats the focuser will be sitting at 232. If I do all my flats at that point, then the flats for the first filter will be 42 steps away from the average focus point for that filter which is 274. Each succesive filter will have flats that are “closer” to the original light focus point.
My suggestion is to capture the focus point data (this may not be possible) for each light frame filter set, average the focus position and set that as the focus position for all that filter’s flats.
I think an alternative is to set up the flats and lights to rotate in such a way that you shoot a flat after each light. But, I don’t have an automated way of putting my flat panel to the scope for that method. And, especially for Narrowband flats, that takes away dark time for imaging. I would like to be able to complete events in sequence, then do my flats based on the actual sequence focuser points used.