37
\$\begingroup\$

Can the milky way actually be photographed like this? I know the image is manipulated, and is probably a composite, but how do you capture such contrast in the milky way? Is an equitorially mounted scope required? Any advice or pointers are welcome, thanks. I'm heading out towards much less light polluted areas middle of next week and I want to try some nighttime sky captures with my 7D.

Additional info: I just found this picture on wikipedia that says: 54s, tripod mounted 5D, 16mm lens, f/2.8, ISO800. What else do I need to be aware of? My 7D is only an APS-C and the widest lens I have is 19mm. I'm thinking that might not be wide enough to capture an interesting shot?

\$\endgroup\$
5
  • 11
    \$\begingroup\$ "Much less light polluted" is the key (and dry helps a lot too). The Milky Way is very dim, so a new moon will also help a lot. And since we're near the edge of the galaxy, it varies with the time of year. \$\endgroup\$
    – user2719
    Commented Mar 28, 2013 at 12:19
  • 1
    \$\begingroup\$ I took Milky Way pictures with my APS-C Canon T1i and a 17-55 f/2.8. No, it's not as fantastic as the picture you link to, but it's still possible! Give it a try. \$\endgroup\$
    – Michael H.
    Commented Mar 28, 2013 at 14:08
  • \$\begingroup\$ @Stan Rogers, it's interesting that you said "dry", because I am heading to the west coast and was thinking of doing it from a beach. Are you saying I should avoid trying to do a seascape then? \$\endgroup\$
    – Octopus
    Commented Mar 28, 2013 at 20:38
  • \$\begingroup\$ @Octopus - It's the local humidity that matters, so it really depends on the weather as much as anything. You can be at the shore and still have dry atmospheric conditions (apart from the occasional direct splash). \$\endgroup\$
    – user2719
    Commented Mar 28, 2013 at 20:52
  • 4
    \$\begingroup\$ Check out this blog post by Luis Argerich: How to Photograph the Milky Way \$\endgroup\$ Commented Mar 29, 2013 at 13:32

7 Answers 7

35
\$\begingroup\$
  1. Use the maximum aperture.

  2. Shutter Speed: Use the 600/(focal length * crop Factor) rule so as to not see the star trails in your picure (Refer here in section 3. Camera settings).
    For your 19mm lens you can go up to 20 seconds.

  3. Highest ISO possible for your camera that you find the images acceptable.

  4. You can use the application: Stellarium to find out if you are in the right time / place to view the milky way.

I did take a picture of the arm of the Milky Way while I was in Chile. Also remember that you need to go far away from any city lights and close to the new moon phase as possible.

Milky Way

\$\endgroup\$
9
  • 5
    \$\begingroup\$ That seems like a really useful formula for minimizing star trails. \$\endgroup\$
    – Octopus
    Commented Mar 28, 2013 at 17:11
  • 1
    \$\begingroup\$ The 600/focal length formula has been around a long time. With cropped digital sensors it had to be amended to 600/apparent focal length. \$\endgroup\$
    – Michael C
    Commented Mar 28, 2013 at 21:01
  • \$\begingroup\$ For crop sensor, just use 400/focal length - Rule of 600 \$\endgroup\$
    – MikeW
    Commented Mar 29, 2013 at 0:40
  • 1
    \$\begingroup\$ Also If you had a choice between a 14mm f2.8 or a 24mm f1.4, I would go with the 24mm 1.4 and you have a 1 stop effective advantage (2 f-stops advantage for 24mm and approximately 1 stop rule of 600 advantage for the 14mm) and stitch the images if necessary. \$\endgroup\$
    – Viv
    Commented Apr 3, 2013 at 23:53
  • 1
    \$\begingroup\$ "Also remember that you need to go far away from any city lights and close to the new moon phase as possible." Considering a night with a First Moon (for example) does the reflected/refracted light after moonset affect astrophotography? Or are you just saying that you should make sure the moon isn't currently overhead? \$\endgroup\$ Commented Oct 29, 2014 at 4:21
16
\$\begingroup\$

Let me answer the question by amalgamating suggestions made by several posters throughout the already provided answers and comments. Hopefully these suggestions taken together will help yield the best possible results.

- Know where the Milky Way is

Obviously, knowing which way to point the camera is important, but actually spotting the Milky Way with your eyes is next to impossible. Know where the Galaxy is in the sky by using recognizable constellations as landmarks. There are programs that can help you familiarize with the sky including Stellarium, Google Sky and for Android Mobiles SkEye, to name a few. There are quite a few available.

- Choose a time and a location

In order to capture the best image there are some important factors to consider. You need to be away from light pollution, so select as remote a location as possible, away from city lights and other significant sources; a cloudless night with no moon. You can Google for moonrise and moonset times. Generally, avoid nights with a full moon. When the moon is waning (after full but before new moon) it will be absent from the sky for the first few hours after sunset, and if it is waxing (after new but before a full moon) you will do best to shoot in the morning hours. Best yet, shoot during a new moon. It won't be in the sky at night.

Also choose arid conditions over moist; directly after rain is a good time. And prefer high altitude locations to lower. The atmosphere is thinner the higher you are and will have less of a distorting effect on the stars.

- Select a lens

You will most likely want to get at least a 40 degree field of view in the frame so you don't want anything more than 50mm for a full frame or around 30mm for a crop sensor. Faster lenses (larger apertures) will allow you to choose shorter shutter times to minimize star trails, and so will wider angles (shorter focal lengths).

- Select the correct exposure

Obviously you want to maximize the amount of light you're working with, but there can be trade-offs for each of the dimensions you have to play with: ISO, shutter speed, and aperture.

First, select your largest aperture, then I would say, choose a shutter speed. Star trails will be a problem when the shutter is held open for more than a few seconds so there is a balance between minimum trails vs maximum light. You can use the rule of 600.

Basically:

shutter speed = 600 / focal length (for full frame sensors or) 
shutter speed = 400 / focal length (for crop sensors)

But the lower, the better for the sharpest image possible.

Then select the ISO that will work with this by using the following formula:

ISO = 6000 * f-stop^2 / shutter

For example:

crop sensor, 15mm lens, at f/4.
shutter = 400 / 15mm (approx. 26s)
ISO = 6000 * 4^2 / 26 (approx. 3692 so choose ISO3200)

At least, that's a good starting point. experiment from there.

- Shoot

Above all, have fun.

Update: - optionally, use post-processing

This will only be possible if you consider yourself a mathematically inclined computer programmer. I recently came across this amazing post from flickr user magnetic lobster, who took multiple 20s photographs and combined them with a mathematical algorithm. This worked incredibly well, even over the bright lights of a big city. Lots of details about his procedure are included.

\$\endgroup\$
3
  • 5
    \$\begingroup\$ I'll have to disagree with one of your points: if it's dark enough to take a good star photo, the location of the Milky Way will be blindingly obvious. It won't be bright enough to see through the viewfinder, but if you need more precision than "point the camera in the right general direction", you're not using a wide enough lens to do things justice. \$\endgroup\$
    – Mark
    Commented May 4, 2014 at 6:58
  • 1
    \$\begingroup\$ @Mark That's all fine and good once you're there. But it doesn't offer much help with regard to exactly when to plan in advance a long journey to a remote location where the Milky Way will be in the sky when the moon isn't, and in a direction that isn't toward a source of light pollution. \$\endgroup\$
    – Michael C
    Commented Oct 20, 2016 at 4:46
  • 3
    \$\begingroup\$ You can use Deep Sky Stacker or similar also for Milkyway pictures. You don't need to be mathematically inclined computer programmer. \$\endgroup\$
    – Grimaldi
    Commented Sep 17, 2017 at 8:27
12
\$\begingroup\$

While all of the answers here offer something helpful, there's one thing that can't be stressed enough: you need to shoot from somewhere free of light pollution. Even if you don't have any major cities around for miles and miles and miles, it's likely that there's still some light pollution cast. You don't realize just how dark outside can be, and how easy it is to see stars, nebula, and the milky way on a moonless night until you're in a spot without light pollution. It's a stark difference.

My own experience suggests that nowhere on the east coast of the US is ever going to be dark enough to shoot successfully, unfortunately. Even in the North Maine Woods where anything is a few hours away -- let alone a big city that might cause such pollution -- isn't actually that dark. I've discovered this on trips to Wyoming, where I was honestly shocked at how bright the night sky was with no pollution. Capturing wisps of the galaxy was downright easy: a 30 sec exposure at f4 and ISO 3200.

Jackson, Wyoming at night

Something interesting about this photo as it relates to this discussion: Jackson is a small town. There are plenty of small businesses and a few big stores, but no skyscrapers, high-rise apartment complexes, or anything like that -- it's a small town. And look at how much light pollution it's casting: you can see it in the photo.

\$\endgroup\$
1
  • \$\begingroup\$ if you're in northern Maine anyway, try hopping the border to New Brunswick. There are a lot of spots that are far enough away from the big villages (Saint John, Frederickton) and there are plenty of roads leading to safe shooting spots that are completely unlit for miles. \$\endgroup\$
    – user2719
    Commented Mar 28, 2013 at 15:34
7
\$\begingroup\$

First of all you have to locate the milky way. For this you may use Stellarium; it calculates/simulates time & coordinates of the milky way.

When you decide the time & location & direction, you need to find a good place where the light pollution is not problem. You'll take long exposures so atmosphere should be as dark as possible. Otherwise you'll see a white-out shot. Cold & clear winter night is much better than summer nights because of air flow. After the rain the weather seems mist/haze free too and stars shine very well. Higher altitudes are better than lowers, atmosphere is much thinner there.

If you use wider angle lens your exposure time can be longer. If you use narrow angle then the time will be shorter to prevent star-trail effect. Of course if you have an equatorially mounted device, trails won't be problem.

Here is a good shutter speeds chart, and more: http://www.bobatkins.com/photography/tutorials/astrophotography.html

At site, try different ISO/aperture/shutter speed settings. Higher ISO levels give more colorful sky (blueish). You may also trigger the speed lite to take pictures of closer things like trees, ground, etc.

\$\endgroup\$
3
\$\begingroup\$

It's entirely possible to capture the Milky Way clearly with just a sufficiently low light camera and a long shutter time. I actually have a photo up in the photo of the week contest on Meta that clearly shows the Milky Way with no alterations (I've made slight brightness and contrast adjustments, but haven't done anything other than full photo adjustments on it.) I used a roughly 60 second exposure from a fixed tripod and I believe a 6400 ISO.

Startracking tripods will help resolve stars more clearly instead of having little lines, but the Milky Way should still be perfectly visible as long as the planet is facing the correct way.

\$\endgroup\$
1
\$\begingroup\$

That photo is not manipulated @Octopus.... You just have to know which direction our planet is facing,.. I have captured milky-way(partially :p) myself once... You have a lot of resources(software and the web .... http://www.astroviewer.com/index.php ) to virtually see constellations and get the direction of the stars. you can go out on a cloudless,lightless night , see using a compass which direction from earth is the constellation ,fix your camera accordingly. And shoot...Make sure u set ISO around 800-1600 to get good contrast.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ Well, it is in a set called 'photo-manipulation' if you look in the right hand margin of that page, although there is no indication of to what extent. \$\endgroup\$
    – Octopus
    Commented Mar 28, 2013 at 6:52
  • 1
    \$\begingroup\$ Says "combination of two images" on the Flickr site. Looks nice, but the Wikipedia photo looks more real, imo. \$\endgroup\$ Commented Mar 28, 2013 at 6:58
  • \$\begingroup\$ The Flickr photo has been blue shifted. The Milkyway would normally have more color in it. \$\endgroup\$
    – AJ Henderson
    Commented Mar 28, 2013 at 13:34
0
\$\begingroup\$

You may also consider the night lens rating of the lens to compare other lenses you might have access to. Using some research and online articles, I wrote this online page which I use to compare various lenses to see which might give me technically best light capturing capability. This is useful to see if I own a few lenses, or even if I plan to rent one, which one of them might be getting in the maximum light in. Of course, that still leaves a numbers of other variables to consider as mentioned by other answers in this post. My lens rating calculation site is at http://www.saurabhkumar.com/projects/astrorating/.

It also hints at the maximum exposure you should use after which trails will begin, and considers full frame vs APSC in the calculation too.

\$\endgroup\$
2
  • \$\begingroup\$ The label for "Max Exposure (mins)" on your calculator is wrong. It indicates that I can shoot 24mm on a full frame camera for 21 minutes before star trails become a problem. 21 seconds sounds about right, though... =) \$\endgroup\$
    – scottbb
    Commented Sep 8, 2016 at 17:19
  • \$\begingroup\$ Oh sorry, bad typo!. Fixed it now. Thanks for the pointer @scottbb! \$\endgroup\$ Commented Sep 16, 2016 at 15:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.