Author Topic: Two questions about flats  (Read 944 times)

Offline CharlesW

  • PixInsight Enthusiast
  • **
  • Posts: 87
    • View Profile
Two questions about flats
« on: 2015 July 27 16:48:15 »
I have a QSI with two download speeds. The fast speed has twice the noise of the quality speed. For the purpose of taking dusk flats, would it be possible to use the fast download speed to get as many flats in as possible, and then use something like TGV Denoise to clean them up? Or, should I just use the quality speed and take fewer flats?

Additionally, is there generally a recommended number of flats to use per filter? I just don't recall seeing a reference for that.

Thanks

Offline Geoff

  • PixInsight Padawan
  • ****
  • Posts: 903
    • View Profile
Re: Two questions about flats
« Reply #1 on: 2015 July 27 17:03:24 »
Hi Charles. Don't ever apply noise reduction to flats.  Part of the reason for taking flats is to correct for pixel to pixel differences in quantum efficiency (PRNU).
As to the other question: I imagine the read noise will be completely swamped by the shot noise in the flats but you should do a quantitative evaluation based on your figures.
See Mike Schuster's post in this thread http://pixinsight.com/forum/index.php?topic=7759.msg51474#msg51474 part-way down as to how many flats are needed
Geoff
Don't panic! (Douglas Adams)
Astrobin page at http://www.astrobin.com/users/Geoff/
Webpage (under construction) http://geoffsastro.smugmug.com/

Offline Stu

  • PixInsight Enthusiast
  • **
  • Posts: 79
    • View Profile
Re: Two questions about flats
« Reply #2 on: 2015 July 28 11:41:32 »
Hi Charles,
I had a discussion once with Kevin at QSI about the number of flats.  He recommended a total ADU of about 1 million. 

So if you look at this link
http://store.qsimaging.com/kb_results.asp?ID=28

And assuming an average flat ADU of 35,000, that would give about 33 individual exposure.  I usually do 35. 

I never use the fast download speed for anything.  And agree with Geoff.  Don't ever process your flats.