Flat Field Analysis

[Aidan]

Having battled with trying to get a good flat field and round stars corner to corner, i think i have finally got to a point where i am happy.  Its been a battle of adaptors, spacers and tweeks many of which were quite challenging.

I managed to test everthing last night and the proof of the pudding was a 10min sub with what i consider to be lovely and flat with round stars all over the image.  

What does still puzzle me is what CCD Inspector was showing me when i ran the initial unguided test subs.   I've attached the 10min sub which shows things are working, but CCDI is still kicking out some strange images when looking at the curvature and 3D plots.    

 

Its more curiosity as to what its telling me... I'm still in the early days of CCDI and learning whats what...  I dont intend on making any more radical changes to my setup unless i really have to!

Rgds

Aidan

 

Quad_600s_M81_Test.fit

[symmetal]

Nice and flat field there Aidan so your spacing is spot on. The error it's displaying is a slight equal tilt in x and y leading to the 44 degree direction shown. I think it incorporates tilt errors as part of the overall 'curvature' error which I believe relates to the variation in FWHM compared to the average FWHM. How it derives the figure of 11.1% I'm not sure. In a previous post you said you had a tilt adjuster in the image train. 

Depending on where the camera rotation adjustment is done in your image train you can narrow down which section is introducing the tilt by taking an image with the camera rotated 90 degrees and seeing if the tilt moves to the other two corners or stays the same. The camera rotator can be a cause of the tilt and you end up with varying tilt errors as you rotate the camera which is a real pain. If you instead leave the camera at a fixed rotation you can use the tilt adjuster, if you can adjust it without removing it from the image train.

Your stars are round all over the field and the tilt errors just cause one corners stars to be a bit larger than the opposite corner, though still round. The error is pretty small and is probably hard to see so it may be worth leaving it as it is and be happy with the result you've got. 

I've just noticed that only 189 stars were used in the CCDI test. This may not be enough to give a reliable indication so it's worth trying to get more usable stars in the image. Pointing at an area of the Milky Way without large area of nebulosity can give you 2000 stars or more. I tend to point around Cassiopeia for my tests. 

Alan

[Aidan]

Thanks Alan,

Ill run it again tonight on the MW and see what happens.  

Overall, i dont think i am going to change anything..     Ill do some rotation tests as well and see how that transpires.  But, visually it looks good and i want to get on and do some imaging after spending the last month testing!

Rgds

Aidan