With pretty much any mirrorless or DSLR camera, a sturdy tripod, fast lens, and just little knowledge, you can now capture landscapes beneath more stars than you ever imagined possible. A camera’s ability to accumulate light allows it to reveal stars far fainter than the naked eye sees; rapidly advancing digital SLR technology now enables usable (low noise) images at the extreme ISOs necessary for star-freezing shutter speeds in very low light.
Before you start
I’m starting with the assumption that you have a relatively new mirrorless camera or digital SLR, one that allows you to capture fairly clean (low noise) images at 3200 ISO or higher. You’ll need to be fairly comfortable with managing the controls in the dark, and know how to get it into manual and bulb modes. For star trails a locking remote release is essential (one that allows you to lock down the shutter rather than forcing you to hold it down for the duration of the exposure).
And of course don’t even think about trying any of this without a rock-solid tripod (you don’t need to spend tons of money, but neither can you assume any tripod will work). A wide (28mm or wider on full frame is best), fast (at least f/2.8, but the faster the better) lens is best. Oh yeah, and take off your polarizer.
Moonlight photography is great for photographing landscapes beneath a few bright stars, but a sky filled with stars (and maybe even the Milky Way) can only happen when there’s no moon and city light washing out the faint stars.
When I go out on a moonless night, whether my goal is pinpoint stars, star trails, or both, I start with a test frame to determine the amount of light my planned image requires. The test frame also allows me to check my exposure, focus, level, and composition in light that’s nearly opaque to my eyes.
My initial test frame is usually no more than a 30-second, high ISO (the goal isn’t a usable image, it’s solely to determine exposure, focus, and composition) and my lens’s widest aperture. After each click I check my composition and focus, adjust, and reshoot. The first frame is mostly to gauge the light; subsequent frames refine both the exposure and composition. I’m usually ready to go after two or three test frames.
Once I have an exposure that works (the desired combination of stars and foreground light), I just need to decide which shutter speed will give me the star effect I want—short for pinpoint stars, long for star trails. With that, finding the ISO and/or f-stop that adds or subtracts the light subtracted or added by my chosen shutter speed is just simple math.
For example, let’s say my test exposure was perfect at ISO 12,800, f/2.8, and 30 seconds. A 30-minute star trail image will gather a lot more light (than my 30-second test exposure), so I start by figuring out how many stops 30 minutes adds to 15 seconds. Since I have to double ¼ minute (15 seconds) seven times to get to 32 minutes, I know going from 15 seconds to 32 minutes adds 7 stops of light. (2×1/4=1/2 minute -> 2×1/2=1 -> 2×1=2 -> 2×2=4 -> 2×4=8 -> 2×8=16 -> 2×16=32.)
A moonless night doesn’t have enough light to see the controls on your camera, the contents of your bag, and the tripod leg you’re about to kick. Needless to say, there’s not enough light to focus either, at least in the traditional ways.
Because we’re usually wide, and very rarely concerned about close detail, all of our night subjects are probably at least 25 feet away with an infinity focus point. Unfortunately, that old prime lens habit of twisting the focus ring to the end for infinity focus doesn’t work on a zoom lens—every focal length has a different focus point (I’ve found this to be true even for lenses labeled parfocal). While I’ve simplified my night photography by usually going with my Sony 24mm f/ 1.4 GM lens, when I do use a zoom (usually my Sony 16-35 f/2.8 GM), I almost always use it at its widest focal length. Not only does a wide lens maximize the amount of sky in my frame, the extra depth of field increases my range of focus tolerance. And sticking with a single focal length reduces the times I need to mess with focus—once I get it sharp, I’m done with the focus hassle.
Despite the hardships, there are a number of methods for focusing at infinity in the dark. Here they are in my order of preference:
1. Autofocus on a bright planet or star. Some camera/lens combinations have excellent autofocus (the faster the lens, the better). I always start by picking out the brightest planet/star. Venus is great, but it won’t be up during the darkest hours of the night. Jupiter, Saturn, and Mars can work, as can Sirius and maybe a few other bright stars. Regardless, you don’t need to know what you’re pointing at—find something bright in the sky, center it in your viewfinder, and try to autofocus. (Any bright, distant object will do—headlights, a plane overhead, whatever.) Don’t forget to take your lens out of autofocus as soon as it’s focused.
2. Live-view focus on a bright planet or star. With my camera on my tripod I center the brightest object in the sky in my viewfinder and lock it in place. I go into live-view mode, center the star/planet in the LV magnification square, then magnify the view to the maximum (it’s 10x on my Canon), and manually focus. Since switching to Sony mirrorless, this is my preferred focus technique and I rarely try 1 or 3.
3. Autofocus on a nearby flashlight. When all else fails, I have somebody stand 50 feet or so away with a flashlight and autofocus on that. If I’m by myself, I rest the flashlight on a rock (or whatever) and walk (stumble?, grope?) 50 feet away. Believe it or not, if I focus my 24mm f/1.4 lens (for example), on a point 50 feet away, I’ll be sharp from about 25 feet to infinity, so you should be fine too unless your lens is significantly longer (which I don’t recommend for night photography) or faster (lucky you). Don’t forget to take your lens out of autofocus as soon as it’s focused.
Don’t forget!: Because there’s no fixed infinity on a zoom lens, if you change your focal length, you must refocus. And no matter what method you choose to focus, you must check the sharpness on the LCD before assuming it’s sharp (once you’ve verified sharpness, you don’t need to refocus or check sharpness again until you change your focal length).
Because I love stars, and it’s the stars that really set a night image apart, most of my night images are at least 2/3 sky. The foreground is usually more of a placeholder, an excuse to dazzle you with the celestial ceiling. But that does not mean the foreground doesn’t matter. Quite the contrary, because the sky is a relative constant, the foreground is the difference between another pretty picture and something that pulls people to a print from across the room.
It’s not necessary, but when possible I always try to include something recognizable, such as the Milky Way (my favorite), or a recognizable constellation like the Big Dipper, Orion, or Cassiopeia. This is especially nice in pinpoint star images. If you don’t know the night sky, spend a little time familiarizing yourself with the major constellations—there are many, many smartphone apps to help with this.
Most people’s vision subconsciously runs along the long edge of an image. Since the primary feature or a night image is the sky, most of my night images are oriented vertically. Regardless of my orientation preference for a particular night shoot, I always make sure I have at least one vertical and horizontally oriented image.
I’m constantly on the lookout for a striking foreground to feature beneath a starry sky. Bold objects without a lot of intricate detail work well, such as a prominent or mountain. Reflective subjects, like water, granite, and sand, work well too.
In Yosemite I like Half Dome for the way it stands out against the sky. For years I struggled getting enough light into the dark hole of the Grand Canyon at night, but today’s digital sensors and fast lenses have changed that. had better luck with Grand Canyon my star trail images because the long shutter time allows enough light at a very clean ISO. My current favorite location for night photography is New Zealand, which I always visit in June (winter). The skies are dark and clear, the nights are long (the Milky Way is up all night in June), and the foregrounds are off the charts
Successful star photography is all about managing star motion—either minimizing their motion or maximizing it. Unfortunately there’s an inverse relationship between the number of stars you capture and your ability to freeze their motion—for any given ISO and f-stop, the longer your shutter is open, the more stars you’ll expose, but the more they’ll move during your exposure.
Pinpoint star images require (relatively) fast shutter speeds to (more or less) freeze the stars’ motion; star trail images us long shutter speeds (either in one frame, or a series of blended frames), the longer the better, to maximize star motion. (Of course it’s not the stars’ motion we’re capturing, it’s Earth’s rotation against a fixed backdrop of stars, but you already knew that.)
Some nights I shoot both pinpoint stars and star trails; other nights I only photograph pinpoint stars. Because a pinpoint star exposure is usually only 15 to 30 seconds, even after I’ve completed my test exposures, they’re the best way to make sure I have everything right before moving on to the quite lengthy star trail exposures.
I’ve seen a formula floating around that’s supposed to ensure pinpoint stars. It’s called the “Rule of 600” (or 500) and says: “Divide 600 by your focal length to ensure a shutter speed that will freeze the stars.” My concern with solutions like this is that they sound far more precise than they are, and they create a false sense of security, often leading to longer or shorter exposures than the scene calls for.
The problem is, the amount of motion is a function of (among other things) a star’s distance from the axis of rotation. For example the North Star, which is less than a degree from Earth’s north axis, will show very little motion in exposures of many minutes or even hours; Betelgeuse, on the other hand, because it’s near the celestial equator will show a significant amount of motion in just a few minutes. For pinpoint stars I think it’s more important to find an exposure that delivers enough light with the least amount of noise.
My biggest problem with exposure speed rules like this is that they can create a worse problem than they correct. Night photography is all about compromise—less than ideal aperture, ISO, and shutter speeds. To me the most unrecoverable compromise, the thing that will render an image unusable more than anything, is too much noise. I generally will forgive the slight amount of star motion of a 30-second exposure (that’s not usually even visible at standard viewing distance) if it saves me from a too dark foreground or unsatisfactory ISO. I find that I’m satisfied with my results if I keep my shutter speeds to 30-seconds and below—the faster the lens, the more likely I am to drop my shutter speed into the 10-20 second range.
I currently (as of September 2019) shoot with a Sony a7SII and Sony 24mm f/1.4 GM lens. I know I can get usable images that clean up nicely with noise reduction software (DxO Prime and/or Topaz DeNoise is my choice) at 12800 ISO, which allows me to stop down to f/2.0 and/or use a 10-second shutter speed. ISO 12800 is higher than I’d use with most cameras, but it seems today’s full frame (and even some APS-C) sensors do fine at ISO 3200, which might require a 30-second shutter speed to get enough light for the foreground.
The Milky Way
The Milky Way may just be the single most beautiful everyday feature of Earth’s night sky. Sadly, increased light pollution has made it all but unknown to the vast majority of us. Once upon a time observing the Milky Way’s glowing band stretching across the sky was for most people a matter of walking out and looking up on a dark, clear night; seeing it now usually requires planning and travel.
As most know, the Milky Way is the galaxy of which our Solar System is a very insignificant piece (the Sun is one star in nearly a half trillion). When you see the Milky Way, you’re looking toward our galaxy’s center and seeing the accumulated light of billions of stars. The dark areas you see aren’t areas without stars, they’re regions of interstellar dust so dense that it obscures all starlight (the occasional pinpoint of starlight in these dark regions are nearby stars between us and the galactic center).
Earth’s position in one of the Milky Way’s spiral arms is kind of like being in the distant suburbs of a large city. While all the discrete stars we view and imagine into constellations are the porch lights of our neighbors (technically they’re part of the Milky Way too, just as some cities have city limits that extend all the way out to the suburbs), when we view the Milky Way we’re looking beyond our neighborhood toward our galaxy’s distant, much more densely populated, urban skyline. Due to our Solar System’s skewed orientation (we don’t orbit the Sun on the same plane on which the Milky Way is laid out), parts of the Milky Way are visible regardless of the side of the Sun Earth is on.
The constellations the Milky Way “passes through” (from our perspective—in reality we’re looking through these constellations to the Milky Way center beyond) include Perseus, Cassiopeia, Lacerta, Cygnus, Aquila, Sagittarius, Ophiuchus and Scorpius, Norma, Circinus, Crux, and Carina. If you want to see it, simply pick one of these constellations, figure out when and where it will be visible (an star chart or app will do), pick a clear, moonless night, and position yourself a location far from city lights. For example, in the Northern Hemisphere Cassiopeia is visible year-round more or less opposite the Big Dipper with Polaris (the North Star) in the center—you might be able to go out tonight to see it (assuming there’s no moon and you can get away from city lights).
But the Milky Way isn’t particularly bright in Cassiopeia—for most photographers (or anyone else who appreciates beauty) it’s the Milky Way center we’re looking for. For that Northern Hemisphere viewers need to look to the southern sky, toward Sagittarius, the constellation that aligns most closely with the Milky Way’s dense (most brilliant) center. And since the Sun is in or near Sagittarius (when we look in the direction of Sagittarius, we’re also looking toward the sun) in winter, we need to wait until Earth has circled around to the other side of the sun—summer.
In other words, viewing (and photographing) the Milky Way’s bright center is a summer (-ish—late spring and early fall will work too) activity. Get out your star chart/app and find a summer night when the moon is below the horizon while Sagittarius is above it (the closer to a new moon, the better your odds). Then get yourself as far from city lights as you can (mountains or desert are great), look to the south, and prepare to be awestruck. Stand there and appreciate the view for a while—when you’re ready to photograph, follow the instructions for pinpoint stars above.
Many people enjoy great success photographing star trails by combining many consecutive, relatively short exposures. In general this approach reduces noise and results in a cleaner image. But since all my images are captured in a single frame (I’m a film shooter with a digital camera), you’ll need to look elsewhere for guidance on that method.
My star trail images are usually 20-30 minute exposures, which I find to be more than adequate to achieve the motion effect I’m looking for. Start with pinpoint star frames and stick with those shots until you’re happy with your composition, exposure, and focus. When you’re ready for star trails, without changing your composition, focal length, or focus:
- Turn on your camera’s long exposure noise reduction (most cameras have it, though it’s usually buried deep in the menu system). LENR isn’t necessary for pinpoint stars (though it may help slightly—results vary with the camera manufacturer), but it makes a noticeable difference in star trail images. The downside of LENR is that it doubles your exposure time because the camera takes a second exposure of the same duration with the shutter closed, compares the results, and subtracts whatever it finds in both images. That means if you take a 30 minute exposure, you’ll need to wait another 30 minutes before viewing your results (which is another reason you want star trails to be at the end of your shoot).
- Put your camera in Bulb mode. On some cameras Bulb mode is one of the choices on the Aperture Priority, Shutter Priority, Manual (and so on) dial; on others Bulb is the step after 30 seconds as you increase the shutter speed.
- Now it’s time to do your exposure math. Assuming you want the same exposure (amount of light) you have in the pinpoint star images, determine how many stops of light your star trail shutter speed will add, then subtract that amount of light with some combination of lower ISO and smaller aperture (larger f-stop number). For example, if your star trail exposure is 30 seconds at ISO 3200 and f/2.8, a 30 minute exposure would add 6 stops (technically a full 6 stops would be all the way to 32 minutes, but those extra two minutes are inconsequential). I usually get my ISO down as far as possible before subtracting light with my f-stop, so in this example I’d probably go with 30-32 minutes, ISO 100, f/4.
- Now you’re ready to shoot. If your camera allows you to block the light entering through the viewfinder, now’s the time to engage that (if you don’t know what I’m talking about, you probably can’t do it, so don’t worry about it). Click the shutter button on your remote, lock it down, and check your watch or set a timer.
- Enjoy the view.
Before I start, let me just say that there are just about as many processing approaches as there are photographers. And there are far fewer absolute right/wrong ways to do things than you might read/hear/see. So what I’ll tell you here is the way I process a night image, rather than the way to process night image. If you already have a workflow you like, or if somebody else tells you a way you like better mine, go for it.
I wouldn’t even consider photographing night scenes in anything but raw. Not only do jpeg captures reduce your margin for error, a jpeg capture makes processing decisions that are difficult to impossible to reverse.
- Cool the color temperature: Since I photograph everything with auto white balance, in my raw processor (Lightroom) the first thing I do with a night image is cool the color temperature to introduce a little blue that gives the scene a more night-like feel. The temperature varies from image to image, but it’s usually in the 3,000-4,000 degrees range.
- Noise reduction: Lightroom/Camera Raw noise reduction is much improved, but I don’t use it as my final noise solution. Rather, I do a subtle de-noise with the Lightroom color and luminosity sliders (you’ll notice much more difference with the luminosity slider than you will with the color slider),
- Clarity: The Clarity slider brings out stars like magic, but you need to be careful about the noise it subtly (insidiously) increases right along with the stars. I’ve found that it’s easy to get so excited by what Clarity does to your stars that you overlook the more subtle damage it does to the noise in the image. I generally magnify my view to 1:1 and slowly pull my Clarity slider to the right, concentrating on the noise and ignoring the stars (as much as I can). I’ll be able to fix a little noise later with my Photoshop de-noise plugin, but I just try to be careful not to create additional problems for myself.
- Dehaze: Like Clarity, the Dehaze slider can make a night image look spectacular, but it’s extremely easy to overdo so be gentle.
- Standard Lightroom processing: While the above bullets are points of particular emphasis, that doesn’t mean that I don’t also apply the rest of my Lightroom workflow to a night image. Exposure, Highlights, Vibrance, Crop, and so on may or may not have their place in any given image.
- Noise reduction (since NR is an art in itself, I won’t go into it in great deal here): I use Topaz DeNoise; it’s the first thing I do when I bring an image into Photoshop. Depending on the rest of frame, I often select the areas most prone to noise (shadows, sky, clouds, etc.) and process them separately from the areas with lots of detail (which may not get an NR treatment at all, or a much gentler treatment that preserves detail).
- Dodge/burn: I find that many night images benefit from subtle dodge/burn brush strokes to smooth tone differences in the sky. For example, I often have to clean up slight vignetting, likely the result of shooting wide open (at an aperture far from the lens’s best). And sometimes I like to moderate the tone difference between the horizon line and the top of the frame. Another problem I occasionally encounter is a subtle brightness on one side of the frame or the other, caused by extraneous light (such as moonlight, nearby artificial light) leaking in from outside the frame.
- Content Aware Fill: The longer the exposure, the greater the chance of something unwelcome finding its way into your frame. Headlights and airplanes are by far the biggest offender. But since the advent of Content Aware Fill, I no longer stress about these things.
- Sharpen: Always my final step, I never sharpen an image until it’s sized for output. Especially with night photography, I selectively sharpen only those areas with important detail—dark shadows are never sharpened. And be careful when sharpening the sky—as with the Clarity slider, sharpening can make the stars pop but at the cost of extra noise. One trick I sometimes do after sharpening is brush with the history slider at around 85 percent (100 percent can sometimes create visible transitions) those areas of the sky without significant stars. And honestly, the more I do this, the less night image sharpening I do and in fact, I often do sharpen them at all.
- Standard Photoshop processing: While the above bullets are points of particular emphasis, that doesn’t mean that I don’t also apply the rest of my Photoshop workflow to a night image.