SGP Focus routine - Focusing on Column Defect


#1

Hi Everybody,

I’ve been slowly learning SGP and implementing it on my system. I have a SBIG STL-11000m, Pyxis Rotator, etc on a Williams Optics Megrez 90 scope, carried around by a Skywatcher Az-EQ6 mount.

Last night when setting an imaging sequence running (Ha session on the Running Chicken) the focus routine kept getting confused, and I’m almost sure was trying to use one of the sensor defect columns in the weighted average for the FWHM calculation, which was giving some weird results. The focus run went to about 220% or something before completing and ran for about 10 minutes (10 Second exposures through a 6nm Astrodon Ha Filter, and about a 30 second download time for a 1:1 binned image).

I snapped a quick screenshot with my iphone but as you can see a weird focus curve - which as the routine progressed se-sawwed a bit. Is there something I can do to calibrate this type of behaviour out? Ie a master dark for the image or something ? tolerance settings, longer focus frame integration time?

Critical focus is paramount to getting a good image, and I’d obviously like to get it sorted.

Any ideas?


#2

Maybe it would be best to use focus offset with Luminance filter instead of auto focusing with Ha filter. Ha filter is too dark and require pretty long exposure times for AF to be successful. Using Luminance filter, more stars would be visible and will not show column defects because stars are much brighter than column defects.

Peter


#3

@jamiep

There is almost no data in your image. The strongest signal is actual your defective columns. In order to correct this, you need to make the star signal stronger than defects. You can do this by:

  • Decreasing your binning to 2x2
  • As Peter mentioned, always use Lum to focus (you need to understand filter focus offsets if you are going to do this)
  • If you don’t want to use LUM to focus (highly recommended), 10 seconds is WAY too short for a 6nm NB filter. You will need more like 60 seconds.

#4

Thanks - I’ll try the longer filter exposures and see how i go.


www.mainsequencesoftware.com