BillT wrote:MikeODay wrote:A deep look at Omega Centauri ( NGC 5139 )
Capture:
9 sets of sub-images with exposure duration for each set doubling ( 1s to 240s ) all at ISO800.
Processing:.
Calibration: master bias, master flat and no darks.
Integration in 9 sets.
HDR combination.
Pixinsight
May 2017
Hi Mike,
Great image. I'm curious as why why you didn't use darks?
Bill
Thanks Bill
I have been into Astrophotography for 3 years now and I have gone through a number of stages with darks.
First I recorded and calibrated with darks because that was the advice on the various sites. However, nothing that I did ( including dithering ) would remove the "rain like" effect it introduced in the dark parts of the image.
Then for around 18 months I used in camera dark subtraction ( ie. long exposure noise reduction ). Again I did this because of the online advice that I needed to subtract darks and it seemed to work in that it did not add artefacts. The problem of course was the imaging time wasted in capturing one dark for every exposure.
Then I read in detail the various blog posts on
http://www.clarkvision.com/articles
My take away from those articles was ( in relation to DSLRs ):
1. Dark subratction does nothing to remove random noise - in fact it adds the random noise from the darks into the integrated sum of the lights
2. Dark subtraction is useful if your camera has significant pattern noise ( mostly older cameras ) or it has a significant problem with hot pixels and you don't want to go to the effort of producing an "hot pixel map"
3. Modern cameras with "dark current suppression" have very little patten noise
4. If the light pollution in your area is significant and you expose so that contribution of light pollution is much larger than the random noise ( and patten noise for that matter ) then you don't have to worry about the noise as it will be overwhelmed by light pollution ( that is, cooling and dark subtraction is only useful if your skies are very dark or for some reason you need to take very short exposures such that the noise is visible through the light pollution)
In my case:
1. My camera is relatively new, has no discernible pattern noise and very few hot pixels ( Nikon D5300 )
2. My skies have moderate light pollution ( pale green zone ). - 240sec exposure at ISO 800 results in the average light pollution level of 0.025 on a 0..1 scale in a linear integrated set of images ( equivalent to the jpeg histogram peak occurring around 30% from the left )
So, I stoped using darks. As a result my images have greatlhy improved. I put this down to:
1. I can now take double the number of lights in the same time ( or in effect double the integration time ); and
2. I am no longer adding random noise via the dark images.