Reader Forums
Astronomy forums are FREE. If you wish to participate you must LOGIN | REGISTER.

Astroimage processing

How to turn a good image into a great one. Ask a question, learn about software, or share your techinques and tips for processing astrophotography.
Flat fielding: the how and why...
Last post 10-08-2008 07:29 PM by adams61. 3 replies.
Sort Posts:
Page 1 of 1 (4 items)
  • 10-03-2008 09:24 AM

    Flat fielding: the how and why...

    I see a lot of talk about flat fielding but I never see any quantitative treatment of it, usually only rules of thumb and frequently wrong information passed off as the truth by the self-assured but ignorant "experts" surrounding us.

    Here's a more detailed treatment of the topic for those of you that want to know what flat fielding is and how it works and it just might offer you some insight to help your images be better:

    http://www.narrowbandimaging.com/flat_fielding_page.htm

  • 10-03-2008 02:56 PM In reply to

    Re: Flat fielding: the how and why...

    I see a lot of detailed information on the Why, but not on the How. Forgive me I'm a newb... Thought I read somewhere that your supposed to point the scope/camera at a white wall and take the images? What if its not perfectly white? Or am I totally off base.

  • 10-03-2008 04:47 PM In reply to

    Re: Flat fielding: the how and why...

    TeleNoob:

    I see a lot of detailed information on the Why, but not on the How. Forgive me I'm a newb... Thought I read somewhere that your supposed to point the scope/camera at a white wall and take the images? What if its not perfectly white? Or am I totally off base.

     

    Fair enough! Sorry I overlooked adding the How....

    1) make sure you have no light leaks if you are shooting twilight flats like i do. make sure you are close to focus too. If you have a first quarter moon, that's an excellent way to focus before the sun sets. I check for light leaks by covering the aperture and shooting a 10-20 second binned exposure> if it doesn't look like a dark, you have a light leak to fix. I use aluminum foil (liberally)

    2) I shoot twilight sky flats. Make sure the shutter transition time is short compared to the exposure time. for my FLI cameras I use about 3 seconds for my minimum flat exposure.

    3) you need to have the sensor exposed to a high enough signal level to be in the FPN limited regime. A good rule of thumb is to expose to about 80 to 90 percent of full well. Typically I am around 80%. You don't want to saturate the flat and you don't want to get into the nonlinear region though.

    4) adjust the exposure time to give you about the same average signal level for each flat. For emission line flats I can shoot them an hour before sunset OK and can get them all done before the light is varying very much between exposures. For broadband filters it isn't so easy: they pass a lot of light so you need to wait later and by then the skies are darkening quickly so I may have to increase the exposure time by 10 to 20% for each exposure. Avoid stars in the flats too....

    5) you can use a white tee shirt tossed over the aperture of a refractor to provide an attenuator/diffuser that will let you start the flat sequence earlier and will also prevent stars from appearing in the flat. I use a rubber band chain to securely hold the tee shirt if I use one: no light leaks permitted around the tee shirt and make sure there are no folds in the cloth or any printing on the shirts.

     after shooting your flat set shoot a set of 20-25 darks of the same average exposure as your flats. I may start at 3.5 seconds and wind up with 20 seconds for my flats so I may shoot darks of 10-12 seconds. The exposure times are short and the darks really don't have much signal beyond the offset so biases are OK too. You just need to remove the offset from each flat before combining them. Again you should try to get almost the same DN (or ADU if you prefer) value in the set, I try to stay around 46,000 DN +/- 1000 DN when I shoot flats.

     then to create the master flat, first combine the darks. I use maxim dl and median combine them WITHOUT normalizing

    Then calibrate the raw flats with that dark master.

     Then you combine the calibrated flats: I use maxim, and median combine WITH normalization turned on.

     

     

  • 10-08-2008 07:29 PM In reply to

    Re: Flat fielding: the how and why...

    lol now that i read this i understand what another amature astronomer was doing when he put a piece of a white sheet and holding it in place with a big rubberband over the end of the tube :) this was at a star party i went to

Page 1 of 1 (4 items)
E-mail Address: Password:
Remember me?

Forgot your password » | Login help »

Not a member? Register » | Why join? »

My Profile

Copyright © 2007 Astronomy.com
Powered by Community Server (Commercial Edition), by Telligent Systems