Author Topic: Flats and PixelMath  (Read 4725 times)

Offline Jules

  • PixInsight Guru
  • ****
  • Posts: 513
Flats and PixelMath
« on: 2012 February 20 07:16:16 »
Hi

I took some sky flats a day or so ago, they looked fine at the time. Now coming to calibrate my images. I find there are faint stars appearing on the sky flats.

I have worked through various permutations of pixel math trying to subtract these stars to no avail. The best idea I had was to create a star mask(inverse) and subtract from the flat. Would clone stamp be a better option?

Has anybody had any experience of this?

Regards

Julian

Offline sleshin

  • PixInsight Old Hand
  • ****
  • Posts: 431
Re: Flats and PixelMath
« Reply #1 on: 2012 February 20 08:39:35 »
Hi Julian,

If you dither your flats during acquisition then the stars will be rejected when you integrate them to make the master.

Steve
Steve Leshin

Stargazer Observatory
Sedona, Arizona

Offline vicent_peris

  • PTeam Member
  • PixInsight Padawan
  • ****
  • Posts: 988
    • http://www.astrofoto.es/
Re: Flats and PixelMath
« Reply #2 on: 2012 February 20 08:47:16 »
Hi,

As Steve says, it's extremely important to dither between flats. To integrate them, I recommend you the percentile clipping rejection algorithm with a very low value for the high percentile.


Regards,
Vicent.

Offline Jules

  • PixInsight Guru
  • ****
  • Posts: 513
Re: Flats and PixelMath
« Reply #3 on: 2012 February 20 11:27:06 »
Steve, Vincent

Thank you for getting back.

I am confused. I do use dithering on image collection. However when it comes to recording sky flats I point my telescope at an arbitrary, homogenous area of sky and start recording changing exposure time to adjust pixel values. I am not even guiding at the time.

Is there an optimum dither in pixels, I use 1 pixel?

Thanks

Julian


Offline mmirot

  • PixInsight Padawan
  • ****
  • Posts: 881
Re: Flats and PixelMath
« Reply #4 on: 2012 February 20 12:29:46 »
Steve, Vincent

Thank you for getting back.

I am confused. I do use dithering on image collection. However when it comes to recording sky flats I point my telescope at an arbitrary, homogenous area of sky and start recording changing exposure time to adjust pixel values. I am not even guiding at the time.

Is there an optimum dither in pixels, I use 1 pixel?

Thanks

Julian

Jules,

You should either turn off the scope ( no tracking - produces star trails) or move/dither the scope position if it is still tracking.
Anything from 50 pixels to a few degrees should work.
Most stars are less than 50 pixels in short exposures. J

ust avoid overlap of stars in the images so rejection can work properly. I find have no problems if I have 4 or more flats using either method

Max


Offline sleshin

  • PixInsight Old Hand
  • ****
  • Posts: 431
Re: Flats and PixelMath
« Reply #5 on: 2012 February 20 12:33:39 »
Julian,

How stars appear in your sky flats depends on whether the mount"s tracking is on or off, not on guiding. With tracking on the stars will be bright points and with it off, there will be dimmer a star trails. I keep tracking on and dither. I'm using CCDAutopilot which automates sky flat acquisition. With tracking on, CCDAutopilot dithers 6 arc minutes between flats. This insures the stars will not be superimposed and are easily rejected when the flats are integrated. For light frame acquisition, I dither 2.5 pixels.

HTH,

Steve
Steve Leshin

Stargazer Observatory
Sedona, Arizona

Offline Jules

  • PixInsight Guru
  • ****
  • Posts: 513
Re: Flats and PixelMath
« Reply #6 on: 2012 February 20 13:29:54 »
Max and Steve

Thanks for the replies, something to think about!

Regards

Julian