Pendleton, OR
ASG Observatory
NGC2244
IC405, IC410, IC417, NGC1931
NGC 3372
NGC 1499
NGC7000
M45
Horsehead & Flame Nebula
Barnard 33
When you use a CMOS camera with deep space imaging, it is very different from fast ISO exposure images taken with many DSLR cameras. Exposure lengths can easily reach 5-10-15 minutes and with that comes inherit heat, signal noise, and light artifacts you may not otherwise see.
Calibration files are designed to eliminate inherent noise in your camera, inherent noise with electronics and heat buildup, and inherent noise from dust and optics. Using a variety of these calibration files can really give your images that cleaned up and boosted look as well.
Lastly, calibration files are very easy to do, can often be done during the daytime, and best of all they are a free way to upgrade your images. There is no downside to NOT creating calibration files other than a bit of unknown or misconception of how or when to use them. Hopefully we can simplify it here.
A quick reference guide to what each calibration file does.
Plan and simple calibration that takes nothing but your camera, lens cap and quick exposures. Bias will take out that noise that every camera just has, dark power-on noise. Just by turning on your camera and taking a very quick .001 second exposure in pure darkness or with lens cap on, you can see noise.
With your lens cap on, these are long ‘exposure’ length images. When you go from a quick .001 second image to a 5 minute exposure, your image will experience different levels of noise. The longer the exposure, the more you can get buildup of dark noise. Many CMOS cameras have amp glow which looks like light leaks coming from the side. This noise differs based on the temperature setting, exposure length, and camera gain.
Flats will calibrate out all the dust particles, lens smudges, and filter imperfections you have as well as vignetting you may have. Vignetting is the change of gradient toward the corners or edges that are often softer in light. Primary reasons for vignetting is small filters used on large camera sensors and light not evenly spread out on surfaces. More importantly Flats simply take your entire optical train imperfections and tries to rid it of these.
Dark flats are used to take out the dark noise that comes in from your flat calibrations. They are independent from your optical train just like Darks. Dark flats are designed to generate that dark noise from taking flats, but I never use these because when taking flats, exposures are typically in the .5 second to 2 second range for most CMOS systems and there really is little to no dark noise generated. Bias will do nearly the same thing here… so they are optional in my opinion, others will beg to differ.
Every camera comes with a certain amount of dead on arrival pixels. Pixels that just don’t work, they never will, they are specs in the picture and can easily be calibrated out and removed with a simple bad pixel map.
This image of LBN 325 was a stack of images and NO calibration files taken into account. Notice the severe gradient over the top, which is more vignetting and flats will take care of that, but also noise and other data errors in this image.
This is the same image loaded with calibration files for Bias, Dark, Flat, and Bad Pixel Map in Astro Pixel Processor. The flats have removed serious gradients and with the use of bias and darks there is no amp glow on the ends and much sharper signal to noise ratio.
How to build a calibration library and be done with them!
No rules here, take quick exposures with the lens cap on in the total dark.
Darks are long exposures, setup can take a couple days. I usually do -20 and for say a 1600 camera, I do gains 0, 75, 139 unity, 200. You will take your most popular exposure lengths. I typically do 15s, 30s, 60s, 90s, 120s, 180s, 240s, 300s, 600s of each gain.
Flats are fairly quick to take, and should be done after every imaging session as dust and imperfections change. In reality, I use a set of master flats for a couple months. Do as you see fit. You should have a flat for every filter if you do narrowband because each filter can have different imperfections.
These are the most optional calibration in my opinion. I don’t use them often, but if you take flats, the idea is to cover your lens and remove the optical train, capture that dark flat noise at the same gain, exposure, temperature. A good bias will do virtually this same thing.
A bad pixel map is a good idea if your system generates one. I know not all stacking software will calculate the bad pixel map, but Astro Pixel Processor is a good choice that does calibrate one for use. Nothing special here is needed, your master calibration process will create one if available
Once you have collected and built all these images, it’s time to load them into your stacking software. ALWAYS use the same software you will stack with to build your master files. You are essentially going to add all the calibration files and then create one master bias, master dark, master flat, bad pixel map for each set of images.
All your calibrations are combined into these masters that you can reuse for the life of your camera (with the exception of Flats, which are good to do once in a while.)
Notice we have here a Bad Pixel Map for each gain, a Master Bias for each gain, a Master Dark for each gain AND exposure length, and then a Master Flat for each optical filter change.
Now when I go to stack images, I can throw in 4 files and I get all that bad signal removed and it didn’t cost anything!