OK, maybe 99.9% is overstating things. Maybe it’s only 99.8%, or 99.7%. You get the idea though.
There’s a lot of information on the Internet in blog posts and videos about astrophotography. In this case I’m taking about what is called wide field astrophotography.
Wide field astrophotography is the type done with a standard camera on a tripod. It includes things like star trails, night sky timelapses and images of the Milky Way. It generally includes some foreground subject for interest and is sometimes referred to as landscape astrophotography or nightscape photography. Some, who are very dedicated to the craft, will use a camera that has been converted for nightscape/astrophotography by removing the IR cut filter on the sensor so the camera is more sensitive to light in the infrared spectrum. This can be done by third-party companies. Some camera makers; most notably Canon, make cameras specifically for astrophotography. These are designated with an ‘a’ in the name (e.g., Canon EOS Ra). There are also cameras that have a cooling system installed to reduce noise from long exposures. Others will use tracking mounts that move the camera in sync with the rotation of the Earth to keep the same scene framed for extended periods so that stacked images align perfectly and no cropping is necessary.
This is different from deep sky astrophotography which istypically done with a camera mounted on a telescope that can see much further into the sky than our wider angle DSLR lenses can, making beautifully colourful images of distant galaxies and nebulae. Tracking mounts, camera cooling and IR-modified cameras are used here as well. A telescope isn’t absolutely necessary. You can do wonderful deep sky astrophotography with a decent telephoto lens
Wide field astrophotography is available to pretty much anyone with a camera, a lens and a tripod. Even phone cameras can produce nightscape images.
The image below was created with a set of DNG files from a Samsung Galaxy Note 9. The phone was mounted on a tripod using a bracket made for phones. Each exposure was 10 seconds (the longest possible on the Note 9) at ISO 800. An interval timer app was used to automate the captures. The total elapsed time was about 25 minutes. You can learn about this and a whole lot more on phone photography in my book The Ultimate (OK, no not really) Guide to Mobile Photography. Yes, the title is intended to be a bit cheeky.
Generally the advice goes that for best results you want a full-frame camera and a wide-angle lens (24mm at most) that has a large maximum aperture (e.g., f2.8 at a minimum).
A full-frame camera isn’t that important. A fairly popular camera among astrophotographers is the EOS 60Da, a crop-frame camera. A very wide-angle lens is nice but not crucial. More importantly you want a camera that allows for manual exposure settings and a more current camera is good since it will tend to have better noise characteristics, beneficial because of the use of higher ISO settings in astrophotography. Today’s cropped frame cameras have vastly better noise performance than full-frame cameras of just a few generations ago owing to technological advancements. A wide-angle lens will let you get more of a scene in view, but that may not be what you want for your specific image.
It’s after that, when we get into the camera settings where things really start to go off the rails.
The advice you generally read, or hear, for night skyscapes is along the lines of: use a high ISO and shoot wide open. Often times the ISO recommended is 3200, or higher. The rationale is that this makes the sensor more sensitive to light and, since there isn’t much light at night you want to get as much of it onto the sensor as possible.
Bullshit point 1: Higher ISO settings do not make the sensor more sensitive to light.
Digital camera sensors have a fixed sensitivity to light and a native ISO setting (often the lowest ISO in the normal range without going into extended settings). When you raise the ISO above that native level what happens is the image is actually underexposed and the computer in the camera compensates for this by brightening the image. This brightening is akin to raising the exposure of an image in editing. It’s a big reason why higher ISO images are noisier. Just like raising the exposure in editing makes noise more visible.
What ISO setting you use is somewhat important. ISO 100 at 30 seconds is going to produce a very different image than ISO 1600 at 30 seconds (assuming the same aperture for both). Many modern cameras have linear (or close to it) sensors. That is for each increase of 1 stop in the ISO, the dynamic range drops by 1 stop due to the increase in noise (the signal-to-noise ratio). What does this mean? It means that you can take that ISO 100/30 second shot, increase the exposure in editing by 4 stops and come up with a result that is virtually identical to the ISO 1600/30 second shot. Now, not all cameras are perfectly linear in response. Most are beyond, about, ISO 400 though. As an aside this is also why Expose to the Right only works at the base ISO setting with most cameras – despite what you may read elsewhere on the Internet; although that one has been fairly well debunked by now. That’s a whole other blog post. Using a mid-range ISO of around 1600 should be fine and will save you from an extra editing step of pushing several hundred ISO 100 images in editing before creating your star trail.
Below are signal-to-noise ratio graphs from the testing sight DXOMark.com for the Nikon Z6 and D800. You can go to the DxO site and see what the response curve for your camera looks like. Those are pretty linear. One caveat about the DxO site: they haven’t tested a Fuji camera in about 8 years. They don’t seem to be able to properly test Fuji XTrans sensors and haven’t been able to figure it out, or haven’t cared to. So, if you shoot Fuji and are looking for information about the response of your sensor, you’re SOL.
Another piece of advice that you’ll see and hear about doing astrophotography is you want to shoot at the widest aperture of your lens (e.g., f/2.8). You want to do this , the reasoning goes, because you want as much light to hit the sensor as possible so that you get as many stars in the sky as possible and don’t eliminate some of the less bright ones by stopping down the lens. This will also allow as much light as possible to illuminate your foreground.
Bullshit point #2: Stopping down the lens has no impact on the number of stars in your image.
Yes, you read that correctly. Whether you use f/2.8, or f/4, or f/5.6 the number of stars in your image will be the same. It also affects your foreground subject matter and we’ll get to that a bit later.
When giving this guideline, it’s probably because either (a) that’s what they’ve read, or been told and/or (b) they’re basing their recommendation on the concept of the Inverse Square Law. I’m guessing it’s (a) because, really, not a lot of people understand the Inverse Square Law.
In very simply terms, the Inverse Square Law says that if you double the distance of your light from your subject you reduce the light intensity to 1/4. You would think doubling the distance would give you half the intensity, right? That’s the square part of it. As an aside, the Inverse Square Law applies to sound as well so you can use it when sizing speakers for an audio system. That’s a completely different blog though, and one that I don’t write for.
The thinking is that the stars are so far away from us that we need to use the widest aperture possible to allow as much of the light from the stars to hit the sensor to compensate for the Inverse Square Law.
The reason the Inverse Square Law isn’t applicable is because of the type of light. The type of light? Yes, the type of light. For purposes of our discussion let’s simplify things and say that there are two types of light: diffuse light sources; and point light sources. WTF are those? Relevant question.
A diffuse light source is one that spreads light equally in all directions. Thing of a perfectly round orb that emits light, hanging freely in the middle of a large space. There is no ability for light emitted from the orb to reflect back into the space. That is a perfectly diffuse light source.
A point source of light is one that emits light waves in a single direction in perfectly parallel lines. A laser is such a light source.
Why does this matter to us? Because the Inverse Square Law only applies to diffuse light sources. It doesn’t apply to point light sources. Given their distance and the fact that we only see them from one angle, stars can be considered, for our purposes, to be point light sources. It’s why, for example, a laser can blind pilots in a plane at 30,000 feet while a flashlight can’t be seen from more than a few hundred feet.
The Inverse Square Law doesn’t even truly apply to things like speed lights, or studio strobes (or handheld flashlights) because those are directed light sources owing to the reflectors and Fresnel lenses used. Even the diffusion panel on a softbox doesn’t fully compensate for the reflectors on the lights and reflective coating inside the the softbox. A light with a shoot-through umbrella would more approximate a diffuse source because the umbrella spreads the light rather than directs it. Despite that, these are still closer to diffuse sources than point sources of light.
What does all this mean for us? Back to Bullshit Point #2 – Stopping down the aperture has no impact on the stars in our photos. Don’t believe me? Of course you don’t. It’s human nature to reject information that is new and different from what we’ve been told, or heard before and believe to be accurate.
Take a look at the two images below. Same scene taken roughly a minute apart. The difference? The first was shot at f/4 and the second at f/5.6.
It’s a bit difficult to tell from these small images, but if you examined them closely you would see that the stars in the sky are no different. What is different is the foreground and the light pollution from the distant town.
Don’t believe me that the sky is no different? Of course you don’t. Here’s the f/5.6 image with a +1 exposure increase in editing.
The sky and stars are the same. What do you also notice? The foreground and light pollution from the town are basically the same as the f/4 image. Recall Bullshit point #1. It applies to aperture, too because of the linearity of camera sensors.
Stopping down will give you the same starry sky and it will reduce the impact of light pollution in your scene. Why does it reduce light pollution? Because that’s more like a diffuse light source.
The advice to shoot wide open is problematic for another reason. Depth of field. Typical advice in astrophotography is to shoot wide open, set focus to infinity and zoom in on the sky to check focus to be sure infinity is really infinity.
Bullshit point #3: Shooting wide open may ignore important foreground elements.
If your foreground has elements that are important and some of those elements are reasonably close to the camera, then you are going to have some of those foreground elements end up not sharply focused due to the concepts of depth of field. Even using a wide angle lens, some elements closer to the camera aren’t going to be considered acceptably sharp since you used a large aperture setting.
You can stop down the lens which will help, although may not eliminate the problem depending on the distance of your foreground elements.
The better way to deal with this is focus stacking and image blending. That’s a whole other blog post. Point being, shooting wide open is not necessarily your best choice.
Shooting wide open can work. If you’re in a place with no light pollution and if you have no foreground elements that are important, or somewhat close to the camera. Having no important foreground elements is plausible. Being in a place with no light pollution is less likely. Not impossible, just less likely. Even in places that are designated as dark sky preserves there can be light pollution.
Shooting at very high ISO settings (i.e., above 1600) isn’t going to make things better because the sensor isn’t more sensitive to light. Plus, if you do it and don’t need to, you’re introducing more visible noise into the image unnecessarily.
This is why 99.9% of astrophotography advice is bullshit 99.9% of the time. OK, maybe 99.8%.
