Thread: Freaky Flats
View Single Post
  #6  
Old 06-03-2011, 05:48 PM
rogerg's Avatar
rogerg (Roger)
Registered User

rogerg is offline
 
Join Date: Feb 2006
Location: Perth, Western Australia
Posts: 4,563
hmm... not sure how much help I'm going to be, apologies if you already know all this, but see if any of this is useful ...

The two flats look quite different. That's odd if they're taking one after the other. It could be just the conversion down to 8bit compressed jpg, but still, they look quite different. The first is generally more even, the second has the brighter area down te bottom corresponding more closely with your end resultimage. The second has big steps of gradient, not a smooth gradient (might be the JPG, but if it's not that'd be a problem). Overall there's a lot of reasons why they could be different but if I took a series of flats and two varied that much when looking at the RAW files I'd throw them away and question what's wrong. Perhaps compare your set and pick a sub-set of flats which look similar. I compare them by stretching the levels but not saving that.

With my ST7 I take flats which have about 25,000 ADU out of 65000, which fits reasonably with your aim of 1/3 down the histogram I think. With my DSLR I just put it on P. This results in a neutral brightness but a lot brighter than yours and probably a bit brighter than my ST7 flats. [edit: I can't remember now, I might under-expose by a stop or two]. They work very effectively. As long as your not clipping data I don't think it matters if they're brighter (just got to make sure they're not over exposed and not under exposed). I would try brighter ones in your situation as you're having troubles.

I don't think you've mentioned what camera/filterwheel you're using. Still would be helpful to know. I would think the reasons would be different if you're doing seperate sets of colour flats for filters.

I think it's clear the flats have been applied too aggresively. They're having too much effect. Not sure if there's a setting you can change for that in DSS or if you should adjust the images before use.

To get the offset of colours you have on stars I'd think you'd need the colours to be offset in the flats. hmm, or perhaps one of the actual light images was accidentily used as a flat and not aligned? that'd produce a bizzar result perhaps like this.

hmm, that's my almost-random thoughts for now...
Reply With Quote