New PixInsight Tutorial: Pre-processing (Calibrating & Stacking) your Images

[kayronjm]

Hi guys,

I will try it out both ways later today to see how they compare. I can post both here for discussion. If it so happens that the order matters, I will happily restructure accordingly.

Best Regards,
Kayron

[topboxman]

Hi Kayron,

I noticed in your tutorial that you use Sony ICX-694 CCD. This is a pretty clean CCD and your Cosmetic Correction and dark subtraction sequence may be irrelevant but it might matter for Kodak CCDs since they are typically noisier than Sony CCDs.

Peter

[kayronjm]

Hi Kayron,

I noticed in your tutorial that you use Sony ICX-694 CCD. This is a pretty clean CCD and your Cosmetic Correction and dark subtraction sequence may be irrelevant but it might matter for Kodak CCDs since they are typically noisier than Sony CCDs.

Peter

Hi Peter,

I do use a Sony ICX694 sensor at the moment, yes, but I used to own an ATIK 383L+ and this had a Kodak KAF-8300 sensor, which can be noisy. The images used in my tutorial were taken with that camera, rather than my current one (QSI 660wsg-8). At this moment in time, I do not bother with darks in the slightest. I just do cosmetic correction and calibrate with a superbias and a master flat - that's it!

[kayronjm]

Hello again,

I pre-processed the tutorial data (Crescent Nebula in HA taken with my noisier Kodak KAF-8300 camera) both ways - the way I describe in the tutorial, step by step, as well as the proposed way of applying calibration prior to cosmetic correction. The following are four small preview boxes of various areas in the images that have lots of hot pixels in the raw exposures:









The image identifiers describe which image is which. The left is my tutorial's method and the right is the method proposed here. I can see no differences, literally. I'm going blind looking for differences!

To look more closely at the above, right-click on them and open the images in new tabs in your web browser. You'll be directed to Photobucket's website. Just click the zoom icon on the top-right corner and the image will load at full size.

P.S. Please forgive the crappy star shapes. This was taken practically when I was starting out in astrophotography and before I knew about drift alignment, or the bad Moonlite focuser misalignment I had with my secondary mirror.

Best Regards,
Kayron

[Geoff]

Hi Kayron. Rather than just eyeballing the images have you tried subtracting one from the other with PixelMath? This will show up any differences.

[kayronjm]

Hi Kayron. Rather than just eyeballing the images have you tried subtracting one from the other with PixelMath? This will show up any differences.

Hi Geoff,

Indeed, I just got round to trying this (I was out imaging last night). In a subtraction, one can see bright pixels scattered around the resulting image. Therefore, I will get round to amending the tutorial ASAP, probably tomorrow. I will be recommending calibration of light frames prior to cosmetic correction. I'm not 100% sure why I chose to write the tutorial in the order that I did. Perhaps it made fine sense to me seeing as I don't bother with dark frames myself. Anyway, the train of thought is now in place so I will make the necessary changes to my tutorial!

Thank you both again for pointing it out and recommending the change.

Kayron

[kayronjm]

I have just finished a big update on the tutorial The changes are as follows:

Practically re-wrote sections 4 and 5 in order to re-arrange how the PixInsight pre-processing workflow is applied. Calibrating light frames is now recommended prior to cosmetic correction. Added note about how master dark subtraction from flat frames during the calibration of flat frames can sometimes result in a PixInsight warning about no correlation. Made amendments throughout the text and on various screenshots to properly reflect the changes made.