How I shot this: Star field
Published on 25 September 2018
So, what’s this?
I am fascinated by the night sky, always have been. I remember reading astronomy books as a boy with pictures by David Malin, not only of nebulae and galaxies, but of infinitely vast colourful fields containing literally millions of stars. The sheer amount of stars in the night sky is staggering and should cause us to stop for a while and think about the things we do down here on Earth. Not that I thought that as a boy, that’s grown-up talk. I just marvelled at the stars, everywhere just stars on those photos, like a jam-packed jar full of luscious gem stones.
To shoot astrophotography on film is a stupid thing, at least so many write. And shooting untracked, leaving Earth’s rotation to wreak havoc on all those nice little pinpointy stars one desires in a wide-field photo, is just a waste of film. Perhaps. It is certainly true that digital photography has simplified astrophotography immensely. But I don’t own a digital camera so I use what I have, which in this case was my trusty Hasselblad 203FE and the fantastic 80 Planar FE.
I am not going to claim that this is a good wide-field astro photo, but I’m pleased with it because the colours of the stars show very well. Incidentally, astrometry.net reports that the region of the sky is around Cygnus. On the original photo each star trails which doesn’t look very nice at all. Here’s the original scan — see the trails? No?
No, that’s right, a small 800x800 pixel image forgives a lot. But enlarge and it quickly becomes ugly. Here are 100% crops of the original 3000dpi scan and the final and edited 3000dpi image.
If only I had a tracking mount for my camera. Added to my wish list for Christmas.
Before I go into the post-processing, I should explain how I exposed this frame. The film used is the wonderful and unforgivingly discontinued Fuji Provia 400X which I pushed one stop to EI800 to suck in as much light as possible without having the film go all crazy on me colour-wise (btw, all my Provia 400X is expired by a number of years). The shutter speed one can use without stars visibly beginning to trail varies with the focal length used. An old table I found online suggested that for 35mm film or “full frame” digital photograpy the times are these:
|Focal length||Shutter speed||Focal length||Shutter speed|
These times are calculated as 800 divided by the focal length. In many cases that is too long and I have seen suggestions online that it is better to divide the figure 500 by the focal length. It all depends. From what I have seen, for wide angle lenses at the wider end, say >20mm, it seems that one can extend the times even beyond what the 800 and 500 “rules” state, but that’s not the case for longer lenses.
|Degrees from celestial equator||Factor|
In any event, the impact of the Earth’s rotation declines with the distance from the celestial equator. This means that one can multiply the time given by mentioned rules by a certain factor depending on how high in the sky the camera is being pointed. I typically follow the list at the right. It should be noted one should use the factor of the area of the sky closest to the celestial equator covered by the lens. For the image above, I pointed the camera around 45 to 60 degrees up. The 80 Planar is not a very wide lens – it’s the approximate equivalent of a 50mm lens for 35mm photography – but to be on the safe side I used the factor for 45 degrees, that is 1,4x, and used 23 seconds. Still, that turned out to be too long, as can be seen on the original scan above. Btw, contrary to what one might thing, longer shutter times won’t result in brighter stars or that fainter stars will become visible; it will only only cause trails. So I could have used a shorter shutter speed and probably simplified the post-processing a bit.
There are really nice star trail photos out there, with long mesmerising trails, resulting from likely very cold nights in the dark. But a star field photo is effectively destroyed by stars that don’t look round and pointy. To counter the trails, I opened the TIFF file in Photoshop (in my case CS6) and duplicated the background layer. On the new layer I changed the blending mode to Darken (this is found on the little pull-down menu activated at the top-right corner of the Layers palette). I then selected the Move tool, zoomed in to 100% and nudged the copied layer using the arrow keys. This way the copied layer, having blending mode set to Darken, will push over the trails on the background layer. One has to resort to a bit of trial and error to find how many nudges are needed. It also depends on the angle of the star trails. If they are more or less vertical or horizontal then it may be sufficient with only a few nudges on one axis. Diagonal trails will need nudges on both axes. As soon as one starts to nudge the copied layer it becomes apparent that one will also lose many stars – compared the crops above. This is just how this method works.
After I had reached sufficiently roundish stars, I turned to the colour processing.
The first stop for all my film scans is Adobe Camera Raw where I will set black and white points and then tweak highlights and shadows a bit. For an image like this, the White and Highligh sliders are very important because they control how bright the stars will appear. I have found that one should be careful with the Black slider because it can easily hide faint stars, and in an image where stars have been ‘lost’ due to the nudging operation mentioned earlier, it is important to try to keep as many stars as possible.
Provia 400X can give a purple or magenta tint to the black, which does not work well for night sky but that is easily removed using the HSL/Grayscale tab’s Saturation function. I often pull these sliders almost all the way to the left to rid the night sky of the tint. But be careful with the magenta slider; too much and the night sky may take on a green tint. Of course, depending on the film used this step may be unnecessary.
The last thing I do is to figure out which colours the stars have simply by looking at the image, and then increase the saturation – slightly – of those colours. This is typically red, orange and blue, and occasionally yellow. Sometimes it helps to also increase the luminosity of the stars’ colours, but one has to be careful as that can turn colourful stars into white dots.
A final note on the luminosity of the stars. It can help to move the Exposure slider a tiny bit to the right to brighten the stars. If done correctly, this will not brighten the dark background. This is image-dependent, though, so it works for some images but not for others.