PixInsight Forum (historical)

PixInsight => Tutorials and Processing Examples => Topic started by: TimScottz8 on 2015 November 05 09:09:06

Title: Image Calibration
Post by: TimScottz8 on 2015 November 05 09:09:06

I've been struggling for a while with Image Calibration. I've tried using the BatchPreproccessing script and I do get a result. But the final image always has a gradient on it which I can get rid of with the DBE, but never precisely. It looked on my images as though the flat calibration was overdoing things and I was winding up with the inverse gradient on my final image.
So I decided to dig deeper and use the Image Calibration tool. As a test I took a calibrated flat image and used that to calibrate an identical calibrated flat image. The result was just the gradient inversion I was seeing on my actual images. Surely I would have expected a totally gradient free image ? What is it that I don't understand about this process ?

Thanks for any enlightenment

Tim

Title: Re: Image Calibration
Post by: mschuster on 2015 November 05 11:02:01

Hi Tim,

Could you put the original .fit/.xisf in dropbox?

I tried this on a master_flat (master bias option off, master dark option off, target and master flat set to the image).

I got a result that is flat: minimum 26288.681, maximum 26288.685.

The result should be equal to the mean of the image.

Thanks,
Mike

Title: Re: Image Calibration
Post by: TimScottz8 on 2015 November 05 13:53:32

Hi Mike

Living in the countryside it took a while to upload to drop box, but here is a link https://www.dropbox.com/sh/9ajvayt6dysi0xu/AACkYp3sGBtjpdSGH-T1lPP4a?dl=0 (https://www.dropbox.com/sh/9ajvayt6dysi0xu/AACkYp3sGBtjpdSGH-T1lPP4a?dl=0)

Thanks

Tim

Title: Re: Image Calibration
Post by: pfile on 2015 November 05 14:35:00

problems like this generally boil down to issues with calibration of the flat subs. if the statistics of the master flat are thrown off by bad calibration, then the scaling factor computed is wrong and the flat overcorrects the lights.

how are you calibrating the flats? what is the duration of the flats? is the camera a DSLR?

rob

Title: Re: Image Calibration
Post by: mschuster on 2015 November 05 14:42:15

Thanks Tim,

I did your flat calibrated with flat test, the result is "flat" (i.e. each channel is a constant across the frame). Here is the result:

https://dl.dropboxusercontent.com/u/109232477/PixInsight/F_5546_ISO800_60s__19C_c_mschuster.xisf (https://dl.dropboxusercontent.com/u/109232477/PixInsight/F_5546_ISO800_60s__19C_c_mschuster.xisf)

I don't know why my result differs from yours. Of course if the flat itself is bad (Rob's idea) this test would not show it.

Thanks,
Mike

Title: Re: Image Calibration
Post by: TimScottz8 on 2015 November 06 01:54:38

Thanks Guys

I think I had been applying darks to the flats inappropriately and this was then generating an offset in the flat I was then using for calibration. I did a manual image calibration last night and it seems to have produced a better result than the batchpreproccessing option. I think I need to revist the batch proccess to see what I'm doing wrong there.

Tim