Kevin_A I have also noticed that the star in my FWHM box never looks solid even when in perfect focus so maybe my guide scopes are all soft…
Unless you have very good seeing, the FWHM is going to be all over the place with exposure times that are shorter than 3 or 4 seconds.
Lets say the spot diagram of your refractor is worse than its Dawes limit (not a bad assumption even with Takahashis). The spot diagram of a 80mm scope can be as small as 8 nm (that is why refractors cost different :-). With a ASI2600 (3.75 nm per pixel), the size of a perfect star (roughly the FWHM) would then be about 2.1 .
With a 80mm/400mm focal length scope, the plate scale is about 2" per pixel. (1.94, but we are just doing back-of-envelope guesstimating) So, translating the star to sky coordinates, would make the star size be around 4".
Decent "seeing" is about 2", poor seeing will be worse. Now, that represents a long term Gaussian profile at 2". In the short term, the variance can jump all over the place.
If you take long exposures, you would expect the seeing to average to 2", and that would bloat your star from the perfect spot diagram of 4" to only sqrt( 4*4 + 2.2 ) = 4.5". But that only holds for long exposures. If you take short exposures, the star could jump around.
Now, this is why multiple stars is crucial, and why also you should not be scared at the FWHM of any particular star is jumping around (if multiple star centroid is working properly).
When you take long exposures, the star simply bloats, but because of the Jointly Gaussian property of two dimension Gaussians, the bloat is perfectly symmetrical. I.e., the star remains round, just bloated. (This is why I snicker when I see people write: you have round stars! The worse the seeing is, the rounder the star would become in a long exposure LOL.
Assuming ergodicity, and all stars used in multi-star centroids have equal weight, 12 stars at 0.5 second exposure is equivalent to 1 star at 6 second exposure.
So, even though each star jumps, they all jump independently, and the centroids average out when you use many of them. Try to adjust your guide camera gain so that you get 12 stars, and the stars that get picked have about the same brightness.
Because the current PHD2 uses SNR as the weight of the stars, you end up with the equivalent of 3 or 4 stars versus 12 stars. I had gathered some images the other night to help me study what the proper weights to use is.
Chen
