Superimposed vs. hyperbolic curve fitting HFD values

[Bret Stott]

Hi Folks,

I’m looking for some advice on how best to proceed with tuning tilt and backfocus using ASCAP as the primary tool. My main issue (I think) is that the run-to-run variation is swamping any trends in the data. One variation in particular seems to be coming from ASTAP — I’m getting different results when I compare the HFD value superimposed on the image vs. the ‘inspector tab’ results for hyperbolic curve fitting and the results don’t (always) seem to track … i.e., while one may show improvement the other could show the same, opposite, or little if any change making it extermely confusing. I’m using an ASI2600MM on an Edge 11 with a Celestron 0.7X focal reducer

Below are two frames with no changes in between taken about 1 minute apart. The top right and bottom left results shows:

image1:
Top-Right: 7.91 (superimposed), 8.005 (hfd 33)
Bot-Left: 9.13 (superimposed), 9.421 (hfd 11)

image2:
Top-Right: 8.01 (superimposed), 8.640 (hfd 33)
Bot-Left: 8.16 (superimposed), 8.494 (hfd 11)

In addition to the absolute values between the superimposed and.the ‘inspector tab’ being notably different, the delta between the frames can also be very different — note how the top-right for superimposed data changed by 0.1 HFD whereas the hyperbolic curve fitting changed by over 0.6 HFD.

Am I misinterpreting the data or perhaps have something setup wrong? Do I need to do some sort of image calibration such as flat frames?

Thank you in advance!
-Bret

[han.k]

Hi Bret,

The HFD routine in the inspector tab uses a larger area for star detection. This allows measuring out of focus stars. But the results should be very similar as the values presented in the viewer . As you indicated they are not. In the next days I will look into the code to find out why, probably fix it and report back.

Han

[Bret Stott]

Hi Han,

Thank you for responding. No rush ... Happy Holidays!

-Bret

[han.k]

Hi Bret,

To get comparable value the option "extra stars" in menu F5 of the viewer should be off. See attached screenshot. Then both only use the stars with a minimum SNR>30.

Still be the reported values will be not the same. The reason is that the tab inspector uses a larger detection window. So the detected stars population will be slightly different since some stars will be missed. In general the FD values of faint stars is slightly underestimated since the flux of the outer areas is partly below noise level.

This skewed result doesn't matter as long it is done for all areas the same.

So in other words it is normal for this routine. I can't fix unless I increase the detection window in the viewer routine which is not desirable.

Clear skies, Han