A Bit About Stellar Photography - and Light Pollution (II)

  Part 1 I Part 2

This is the second and final part of the tutorial which started here.


In Part One I tried to show why it is a good idea to try out the highest possible aperture and ISO, when one wants to capture as many and as weak stars as possible. I further tried to demonstrate how to minimize both noise and light pollution with simple means.


Now let us see how far these fairly simple techniques may actually take us.


1. Choice of Celestial Object and Photographic Lens

First, let us take a look at the original “raw”, unprocessed file from Part 1 (which shows the constellation Hercules high in the sky in a westerly direction at around 2 o'clock at night here in early September. Isn’t there something "woolly" about one of the stars?

Unfortunately, this is not too evident in an image that has been compressed from 3872 to 1100 pixels in width, but an uncompressed and post-processed crop immediately confirms the suspicion. This is no ordinary star. It clearly has a kind of nebulous "surface":

In order to get a closer look at what this might be, we need to have a more powerful lens. We shall need a lens with a focal length of about 135 to 200 mm, and we must have a reasonable resolution of our lens too. The resolution depends directly on the absolute, physical size of the aperture. Thus, not only for the sake of light capturing power but also for detail's sake, we need to have a both reasonably long-focal and reasonably fast lens. So, I choose my old, manual Tamron 200 mm f/3.5 (model 04B from 1979).


2. Additional Equipment
In the picture above we see that already at 55mm focal length and an exposure time of 15 seconds stars are stretched out in small lines (or rather arcs) caused by earth's rotation around its axis. With a 200mm focal length, these arcs would become 4 times longer with the same exposure time so, now there is no way around: I must have camera follow the stars’ apparent motion across the sky:


I therefore bought a used - so-called equatorial - mount, which cost me some 200 EURO. Should this be considered as "extravagant" or "exotic"? I do not think so, because other photographers happily spend some 250+ EURO on a Manfrotto tripod + ball head plus around. 400 EURO for, say, a Canon EF 70-200 mm f/4.


The gear that you see in the picture is (besides the Pentax K200D camera body) as follows:

  • Used Meade Autostar fitted with computer control – EURO 200. You may get a brand new mount with motor on one axis and without computer control at approx. the same price.

  • Tamron 200 mm f/3.5 Model 04B telephoto lens - bought used on e-bay for approx. GBP 50 including a PKA adaptall-2 mount, which often sells alone for the same or even a slightly higher price. The lens alone would probably cost around 30-40 GBP.

  • Walimex right angle viewfinder – 65 EURO. Maybe not the finest (plastic) quality, but very fine for the price and its use - for me almost a must when I have to set and check the direction and focus (on a suitably bright star) with the camera in an awkward position.

  • Finder scope consisting of a Tamron 80-210 mm. f/3.8-4 (Model 103A) purchased on QXL for 7 EURO and a Tamron "Tele View Adapter" 20mm right angle eyepiece, which I have owned since 1982. A used spotting scope of a reasonable quality would also do the job (for example recently seen advertised with a price tag of less than 60 EURO.

 All in all we are talking extra gear that has cost me about the same as what you pay from new for a really good photo tripod. And less (: equatorial mount + a reasonably fast telephoto lens) may do for a start.

3. Set-up and First Shots
Now we are ready; the equatorial mount has two axes of which the axis of rotation (the polar- or right ascension axis) must be parallel to the earth's axis and thus, pointed toward the north celestial pole. So, I align it such that it points toward Polaris / the North Star, which is adequate for today's purposes - but not for exposure times much more than 15-30 seconds when I use a 200 mm lens. I have already aligned camera and finder scope so that they point at the same, remote point. (This can be one in daylight). Now then, it is "just" to switch on the polar axis drive and go.

Using the finder scope and turning both axes (the other axis is called the declination axis) I find my target - which can easily cause some frustration as long as you are inexperienced - and check the camera's focus and alignment (provided that I can see my target in the viewfinder - otherwise I will have to start at a bright star and then move from there towards my target). And my target today is that fuzzy "star" which we will name M13 (“Messier 13”), and once placed in the field of view, where I want it, I take five identical exposures at f/3.5 (full aperture), ISO 1600, and 15 seconds of exposure time. Additionally, I take an "LPCN-frame" as discussed in Part 1


Next, I could now subtract my LPCN frame and go through all the steps from part 1, but today I would rather try out with:

4. Stacking in Registax
It would lead us too far to go through all the mysteries of stacking here. The program itself and the manual can be downloaded (for free) here:


Why use stacking in astrophotography, where all objects are infinitely far away, and we therefore do not have problems with depth of field? Well, because of air turbulence light in its passage through the atmosphere is constantly being refracted and bent, and even on the best of nights, the image of a planet, a star or a detail on the moon will constantly dance a bit up and down, forth and back and thus, smudge the image a little. This stacking can remedy by averaging and recalculating the image over many exposures. (Usually many more than the 5 that I use here). This applies right from the relatively fast shutter speeds, about 1 / 50 s or so and longer. FURTHERMORE – and that is a great value to - this averaging helps smoothing the colour and luminance noise (but not pixel noise that will still show up as bright points in the final image). We see the result of my stacking here:


Even with only 5 images in the stack the weakest stars become somewhat better defined and the noise has been significantly leveled.


With this stacked image, we can proceed with processing such as I described in Part 1. Therefore I will just show the results of the intermediate steps in brief succession:

5. Contrast improvement in CURVES

- exactly as in Part 1.

6. Possible post-focusing with FOCUS function

You can, if you will, "improve" a little further on the sharpness increases that we already got in the stacking process.But beware, that you do not exaggerate! The stars will rapidly pixelize and loose their last degree of "naturalness".

7. Fine-tuning i LEVELS

- completely as in Part 1

8. Done!

again, this compressed image does not fully show what we have achieved, but here's a close-up from the original compared to the stacked and finished image:


plus an uncompressed 100% crop "panorama" of the finished image:

- taken at 5 x 15 seconds in the center of town with an old 200 mm telephoto lens from the seventies-eighties. We actually also see stars down to 12th magnitude, which is approx. 160 times weaker than the average human eye can see under a good, dark night sky (magnitude 6.5), and approx. 1600 times fainter than the stars (magnitude 4) that I can personally see with the naked eye from my balcony in the heart of the light polluted city.

A little final "statistics" for nerds (like myself)
M13 takes up quite a lot of space in the sky. Here we only see the inner approx. 20% of this globular cluster. If it was equally bright from center to the edge we would see - under a dark sky - a faintly glowing disc with a diameter of approx. 70% of the full moon, (23 arc minutes)


It is also quite large in reality. The light that began its journey from center of the cluster when my Tamron was built in 1979, has now travelled about half the distance to the rim. (M13’s diameter is 145 light years).


It contains many stars. In the picture we see the combined light from around 100,000 stars. (The scholars are somewhat divided, but some believe that M13 contains a total of up to 1 million larger and smaller stars combined. The number should at least be counted in hundreds of thousands).


It belongs to our own Milky Way galaxy, but it is quite far away. When the ice after the last Ice Age melted, and the first humans arrived in what is now called Denmark, the light than is now depicted in this little tutorial, had reached just halfway from home on its voyage to my balcony and my CCD sensor. (The distance is 25,100 light years).

  Part 1 I Part 2





Copyright © 2010 - Steen G. Bruun