Pixel diameter of stars

[CharlesW]

I heard Tony Hallas speak yesterday and he suggested dithering a distance of two star diameters. Is there a method in PI for measuring a particular star's diameter in pixels?

[Carlos Milovic]

You may use the FWHM or the standar deviation to get an estimation of the star size. If it is a gaussian distribution, then a diameter of 4 standard deviations includes 95% of the light. 6 times, 99.7%.
Anyway, if you want to make a good dither, I would go for nearly a dozen pixels, to make a good uncorrelation of the pixels at the integration stage.

[Juan Conejero]

The tool to measure star sizes and shapes is DynamicPSF in PixInsight. There is also the FWHMEccentricity script, that automates DynamicPSF on a large number of automatically detected stars. Note that all of the stars, with the exception of very dim and saturated stars, have approximately the same dimensions and shape in a linear image (neglecting optical aberrations). This is known as the point spread function (PSF) of the image.

However, as Carlos has said, 10 pixels or so are usually enough for dithering.

[CharlesW]

Thank you gents. Appreciate the help.

[jerryyyyy]

BTW all this has been very helpful in my various collimation adventures with my Takahashi 180 ED. 

Too bad you cannot do a real time/live view calculation like CCDInspector.  I find your post-hoc analyses more useful than CCDInspector's, but they can do Live View.