37
\$\begingroup\$

Many photos of Milky-way show it in nice colors instead of just the black of sky and white of the stars. Is the key to having those colors in the capturing of the image or in the post processing, or both? Or am I just doing it in a wrong time of year when the visible part of Milky-way is simply not showing color?

Here is a sample photo of how Milky-way looks like in Finland in September:

image

APS-C sensor camera, 18 mm, 30 sec, f/3.5, ISO 3200 - captured in pitch-black darkness of Finnish countryside, far far away from any citylights, at time when moon was still below visible horizon.

What I'd like to achieve is something like the photo in Vivek's answer to How do I capture the Milky-way?

I've just bought a wireless shutter release for longer bulb exposures, and have plans to build a barn-door tracking mount. But is longer exposure the needed trick, or what is?

\$\endgroup\$
4
  • 2
    \$\begingroup\$ Different parts of it appear to be much more or less colorful than others. Take a look at this image of it that allows you to freely move around it: galaxy.phy.cmich.edu/~axel/mwpan2/krpano \$\endgroup\$
    – dpollitt
    Commented Sep 29, 2013 at 23:04
  • 4
    \$\begingroup\$ All other things being equal, longer exposures will make the stars appear whiter as more and more of them blow out all three color channels. \$\endgroup\$
    – Michael C
    Commented Sep 29, 2013 at 23:34
  • \$\begingroup\$ @dpollitt - I just got to view the rotating Milky-way panorama and it sure is not very exiting section of the Milky-way there near Andromeda. Not the dullest part either, but it really is nothing compared to the brightest section. I guess I'm trying this during wrong season. Thanks for the link, it was helpful. \$\endgroup\$ Commented Sep 30, 2013 at 4:14
  • \$\begingroup\$ And also use the "daylight" WB setting. \$\endgroup\$
    – FarO
    Commented Oct 13, 2015 at 11:17

6 Answers 6

47
\$\begingroup\$

The vast majority of night sky photos have been boosted in post to achieve their brightness. This is more true for cameras with smaller sensors than for cameras with larger sensors, but in general, even if you shoot the night sky at ISO 3200, you are going to need to boost exposure to get one of those nice, bright single-frame Milky Way shots.

There are a few things you can do to increase the brightness of your night sky shots.

First and foremost, don't be afraid to push ISO. I own a Canon 7D, not particularly great at high ISO, and I usually use ISO 1600 and 3200 for my night skies. If you are using ISO 100, 200, or 400, your too low (unless you are also using an f/1.4 or faster lens, and even then, I would still recommend using ISO 1600 at least!)

Second, find the darkest skies you possibly can. This can often be difficult, especially in densely populated areas. As an example, almost the entire eastern half of the United States is riddled with light pollution, as can be seen by this view of Dark Sky Finder. On this map, blue and green are acceptable for night sky photography, but still prone to issues with light pollution. In my milky way shot here, I was 50 miles from Denver, in a blue zone, and I still had considerable problems with the metro area light bubble over the horizon (note, this image is heavily processed, original below):

enter image description here

The darker the skies, the brighter night sky objects can be made (key point here..."made"...there is a post processing component that I'll discuss later.) An additional note here...another key factor in achieving dark skies is finding moonless skies. New moons are the best time, and you usually have several days on either side of a new moon where you basically have no moon at all (it either rises or sets with the sun). Even when the moon is out, so long as you time your trip during a period when the moon is not actually in the sky should be fine. That usually means later (i.e. some time after midnight) for waxing moons, and earlier for waning moons. Around full moon, night sky photography is usually out, unless you want the moon to be involved for some reason.

Third, use the longest exposure you can. In the past, this has largely been guided by the 600 Rule, which says divide 600 by the effective focal length of the lens to determine the longest exposure that will not produce star trailing. Today, with ever smaller pixels, the 500 Rule seems more effective. Here is a table for some common wide-field focal lengths:

 Focal Length |   FF   | APS-C (1.5x) | APS-C (1.6x) 
=====================================================
    10mm      |   --   |     33s      |     31s
    14mm      |   35s  |     23s      |     22s
    16mm      |   31s  |     21s      |     19s
    18mm      |   27s  |     18s      |     17s
    24mm      |   20s  |     14s      |     13s
    35mm      |   14s  |     10s      |      9s
    50mm      |   10s  |      6s      |      6s

In general, wider lenses allow longer exposures, as the angular movement of the sky covers a smaller portion of a wider frame for any given unit time. On APS-C, a 10mm lens will offer the longest exposure time, while on FF a 14/16mm lens will offer the longest exposure time. (Note that 16mm on FF is the same as 10mm on APS-C, so the option of a 14mm lens on FF is an added benefit.) Also don't rule out fish-eye lenses, which offer a full 180° field of view, and therefor an even longer potential exposure time, albeit with a warped projection. It should also be noted that, while a 50mm lens in the table above allows only a very short exposure, it is also often possible to get a 50mm lens with a one-stop, sometimes an even two-stop advantage over other lenses. The 50mm f/1.8 lens is usually the cheapest lens in a lineup, and finding a 50mm f/1.4 lens is often not that hard. That is like exposing for one or two stops longer with any other lens, so the short exposure time is often still workable with a 50mm lens.

It should be noted that the 500/600 Rule assumes identical output magnification. With night sky photography, that is a pretty good assumption, but not necessarily always true. If you intend to crop for any reason (i.e. to blow up a nebula or galaxy), you should be applying your additional crop factor as well. Using a longer focal length is usually a better alternative, however longer focal lengths quickly run into exposure length issues anyway without further measures.

Fourth, if you have the option, get a camera with the biggest pixels and lowest high-ISO read noise you can get your hands on. Technically speaking, the Canon 1D X would be the best astrophotography camera on the market today. The Canon 5D III is a far more accessible alternative, and still offers pixels considerably larger than any APS-C part on the market. Larger pixels not only increase the amount of exposure time you have before star trails start occurring, but they gather more light in total during that time as well, so its a double benefit.

Fifth, photograph on nights that have good seeing, low atmospheric distortion, etc. The amount of stellar light that reaches an itty bitty camera on the surface of the earth is often dictated by how much of that light is scattered by the atmosphere. Even outside of urban light pollution bubbles, skies full of dust or moisture will warp and scatter a lot of the light reaching the atmosphere. The clearer and crisper the skies, the better your exposure will be. There are various sites on the internet that can probably help you find clear, dark skies with good seeing.

Finally, remember to post process. Even an ISO 3200 shot under decently dark skies is going to be fairly dim at the maximum 500 Rule exposure time. A sky devoid of dust or significant moisture, with good seeing, will produce some beautiful exposures. Mountainous regions, particularly above 11,000 feet, offer this kind of sky in spades, however are less accessible. For any other area, including blue areas on the Dark Sky Finder site, your night sky photos will require some exposure boost and tone mapping in post to fully bring out the detail you are looking for. As an example of how extreme edits may need to be, here is the original version of my shot above...still riddled with light pollution from a city fifty miles away:

enter image description here

Despite the sky clarity problems, you can bring out a lot of detail and color with some processing. You will usually end up with very bright stars, which can be a bit unappealing, and the only way to really fix that is to find darker skies.

One final option, for those who have the money, is to get a tracking mount. Good telescopes usually come with an equatorial tracking mount which will nicely track with the sky. This is only really an option if you are just photographing the sky...any included landscape/foreground will blur as the camera tracks. If you buy a tracking mount or better yet a decent telescope, that opens the door to deep sky astrophotography, which is complimentary to the kind of wide-field astrophotography I've discussed so far.

Photo Statistics

Camera

  • Canon EOS 7D
  • Canon EF 16-35mm f/2.8 L II
  • Gitzo GT3532LS Tripod
  • Gitzo GH1780QR Ball Head

Exposure

  • Focal Length: 16mm
  • Shutter: 25s
  • Aperture: f/2.8
  • ISO: 3200

Processing

  • White Balance: 3590 (5250)
  • Tone:
    • Contrast: +45
    • Highlights: -100
    • Shadows: -30
    • Whites: +71
    • Blacks: -25
  • Presence:
    • Clarity: +35
  • Tone Curve:
    • Highlights: +20
    • Lights: +20
    • Darks: -25
    • Shadows: -60
  • HSL/Color/B&W:
    • Saturation:
      • Red: -35
      • Orange: -16
      • Yellow: -5
      • Aqua: +20
      • Blue: +45
      • Purple: -5
      • Magenta: -35
    • Luminance:
      • Red: -65
      • Orange: -20
      • Blue: +70
      • Purple: +100
      • Magenta: +35
  • Sharpening:
    • Amount: +40
    • Radius: 0.5
    • Detail: 20 (25)
    • Masking: 70 [Causes blurring of smooth areas, which is what I wanted]
  • Noise Reduction:
    • Luminance: 80

(Note: Original values in parentheses when difference is important.)

\$\endgroup\$
19
  • 3
    \$\begingroup\$ Impressive answer, thank you for taking the time to provide examples and the great text as well. \$\endgroup\$
    – dpollitt
    Commented Oct 7, 2013 at 22:41
  • 1
    \$\begingroup\$ @MichaelClark: I think thats pretty much exactly what I said...no? \$\endgroup\$
    – jrista
    Commented Oct 8, 2013 at 7:26
  • 1
    \$\begingroup\$ @EsaPaulasto: Our eyes can often see the milky way better than a camera. In my sample photo, I couldn't see any light pollution with my eyes...only a slight glow behind the trees, but as far as my eyes were concerned, the sky was black and the milky way stood out better than in the photo. Cameras see differently than we do, because our eyes are really more of a combination of camera+post processing software (the brain). That said, if you are getting truly black skies, you are just simply not exposing long enough...expose longer, and at a higher ISO, and with a wider aperture. \$\endgroup\$
    – jrista
    Commented Oct 8, 2013 at 19:03
  • 1
    \$\begingroup\$ @drfrogsplat: It really doesn't matter the generation of the sensor. Any technology that can be applied to smaller pixels can be applied to larger pixels, so there will NEVER be an SNR advantage to smaller pixels in the grand scheme of things. Greater area per pixel, stronger signal, greater SNR, lower noise. That's just the fact of things, at least when it comes to astrophotography. Regular photography may be different, as what often matters most is total light per sensor area, rather than per pixel area. In AP, the total light per pixel is really the most important thing. \$\endgroup\$
    – jrista
    Commented Feb 27, 2014 at 0:25
  • 1
    \$\begingroup\$ @MichaelClark: It depends on what you want and how you stretch. Are you more interested in seeing the foreground stars, or more interested in seeing the milky way itself? If you expose for the stars and try not to clip any of them, then your background sky SNR will be woefully inadequate. This is a fundamental dynamic range issue with digital cameras...we don't have even remotely close to enough DR to preserve all the foreground stars, and get decent SNR on background objects or the milky way. You have to choose. Stars can be managed, reduced, enhanced, if necessary. \$\endgroup\$
    – jrista
    Commented Oct 19, 2016 at 22:32
13
\$\begingroup\$

Time of year plays a big part in capturing a nice view into Milky way. So does the location on Earth where you are with your camera.

It appears the sample photo (in the question) above is of a not-so-exciting part of Milky-way. The widest and most colorful part of Milky-way is where the direction is towards the center of our galaxy. Because earth's axis is tilted, and the planet earth travels around the sun, this direction changes, not only by time of day, but also by time of year. Unfortunately the center of our galaxy is not well in view for photographing it in Finland (North of 60° latitude) at any time of year, except in daytime during summer, which is impossible time for photographing it.

Here is two screenshots of different sections of Milky-way:

Milky way with Andromeda in sight
^^ This is the same view as in the sample photo in the question.

Milky way center mass
^^ Here is a wider and brighter section of Milky-way.

Easy to see how time of year/day makes a big difference in which part of the Milky-way is visible. A nice photo of the center mass of our galaxy is shown in another question "What kind of camera.." photographed by Bala Sivakumar. AFAIK that view is not possible to photograph in Finland, and I don't mean the building in the photo.

You should use Stellarium software to find a good time for photography. First take a look at a page suggested by @dpollitt: galaxy.phy.cmich.edu/~axel/mwpan2/krpano and find a view you'd like to try to capture. When found, zoom in to see the star/galaxy designators on the image. Pick one of them.

In my case I chose to focus on M16, the Eagle Nebula, so I put that into the search field in Stellarium and set it to keep M16 centered on screen. A focus object is needed because we can't tell the program to "keep the nice part of Milky-way in center". Then I started playing with date and time values.

Stellarium screenshot
^^ A screenshot of Stellarium software.

This way I found out that to catch a good view on the wider portion of Milky-way I should wait till the week after easter 2014, and get out with a camera each morning between 3 and 4 a.m. For the screenshot I adjusted Stellarium settings to increase the brightness value of Milky-way and lower the Light-pollution level.

The rest of the actual photographing of colorful Milky-way goes as suggested in other answers to this question, and generally as in about every question/answer concerning Milky-way photography.

\$\endgroup\$
7
\$\begingroup\$

There are several things you can try.

  • A full frame sensor will allow you to saturate the colors more in post processing before noise becomes an issue. If it has larger pixels (which most do) it will also allow higher exposure before color is lost to full saturation in all three channels (more on that below).
  • Stacking multiple images will also allow you to increase saturation because stacking images also reduces random noise.
  • Use the formula in this answer to calculate your ISO setting. In your example above it calculates to about ISO 2500. The higher the exposure, the whiter the stars will appear. Even if a star is, for example, more green than red or blue, if you blow out all three color channels it will appear white.
  • Set White Balance using color temperature and experiment with different values to bring out cooler or warmer colors.
  • Shoot from the darkest sky possible (away from light pollution), at the highest altitude possible, through the driest air possible, when the Moon is not in the sky.
\$\endgroup\$
5
  • 1
    \$\begingroup\$ Thanks. Can't get a ff-camera, so that leaves me with stacking at lower ISO. You suggested 5200 K in your answer to How to set White balance in a photo of stars? and it was a good starting point, I did not move far from it. \$\endgroup\$ Commented Sep 30, 2013 at 0:01
  • \$\begingroup\$ 5200K is a color temperature, not an ISO. That aside, the other question made no mention of maximizing color saturation in astrophotography. \$\endgroup\$
    – Michael C
    Commented Sep 30, 2013 at 0:53
  • 1
    \$\begingroup\$ Hehe, of course 5200 K is not ISO, K stands for Kelvin and measures temperature, in this case color temp. And that's what the other question was all about, nothing more to it. Your answer here is very enlightening, and I will do a stacking job at lower ISO and give it more saturation. My blowing all color channels seems the likely reason to missing colors, if there ever was those colors to begin with. \$\endgroup\$ Commented Sep 30, 2013 at 1:19
  • 1
    \$\begingroup\$ a full frame sensor isn't inherently less prone to noise. \$\endgroup\$
    – Agos
    Commented Sep 30, 2013 at 9:14
  • 2
    \$\begingroup\$ It is if the amount of light falling on the lens from the field of view projecting on the lens is the same and the pixels are significantly larger than those on the 7D, which is true of all Canon FF cameras on the market at this time. The 7D's pixels are 4.3µm wide. The 1D X, 5D3, and 6D respectively are 6.9µm, 6.25µm, and 6.54µm wide. Thus each pixel on the FF models has over twice the surface area with which to catch photons per pixel as the 7D does. \$\endgroup\$
    – Michael C
    Commented Sep 30, 2013 at 20:49
4
\$\begingroup\$

Esa Hi. I too, after taking a not-dissimilar picture to your first effort, had similar thoughts and determined to see if I could improve. I'm still on that "journey", loving every minute of it, and I think improving. Funnily enough I am half Finnish, I have family in Lahti, Helsinki, Turku and Kuusamo, and have a brother called Esa.

Some good stuff in the existing replies. Initially as a newbie I just linked to my blog, but I was told that's not acceptable, so here's an extract:

(link to my full blog here http://www.slidingseat.net/stars/stars.html#startingout )

Taking My First Nightscape Pictures:

from early attempt to better attempt

I often visit Baltimore in South West Cork, about as SW as you can get in Ireland. Walking back from the pub one November night, I was SHOCKED by the sheer blackness of the night sky and the dazzling array of stars. Constellations were not even recognizable, drowned out in the sea of stars. My interest in astronomy suddenly resurfaced.

Round 1 – a Horrid Mess

As of summer 2016 I hadn't really seen the Milky Way for donkey’s years! Out there, the Milky Way is easily visible to the naked eye even straight after leaving a bright room. One evening looking up I thought “I wonder if I can photograph that?”. Being a keen photographer, I reckoned I ‘knew a bit’ about photography – it turned out “you know nothing, Jon Snow”.

Nonetheless I quickly retrieved my camera, attached my fastest wide zoom lens, plonked it onto a tripod and pointed it up. Only then did I think: “Er, what settings to use?”. Obviously the widest aperture and its widest angle (24mm f/2.8) and ISO 3200 (because it seemed "about right"). I chose 30 seconds exposure, but was aware that stars might streak, as the earth rotates noticeably over even as little as a half-minute.

The result, from early August 2016, is here, my first Milky Way image, looking up SSW at around 1am.

early attempt

It’s a horrid mess of a picture. Yes, you can see the Milky Way, but that’s about it: out of focus, no other context, no colour, heavily streaked stars, noisy, boosted to hell in Photoshop. Funnily enough, for a while I was quite impressed, though I didn’t really solicit opinions. If you've never photographed a night sky before, you too may be impressed, but this is really not very good.

Round 2 – 8.5/10 for Composition, 3/10 for Execution

Five months later, Christmas 2016, I had another go. It's obviously a slightly different shot, but it was taken from the same patio as above. I tried to get the Milky Way and the Andromeda Galaxy visible in a recognizably local setting.

Andromeda & MW

I used the same set-up as Round 1, but shorter exposure and lower ISO: 30s at f/2.8 and ISO 1600. I showed this picture around and received many compliments, it’s undoubtedly more interesting. Personally I like it, but only as a small image. Blown up, it’s unpleasantly grainy and the star-trailing is evident, especially top-left (click the picture and a full-size image will open, it'll be obvious). To make the Milky Way stand out, I had to really “stretch” the faint detail, exacerbating the noise.

Describing the scene: The bump towards the left with two bright lights is Mount Gabriel, atop which sits Europe’s western-most air traffic control radar domes. The brightness at far left is the town of Schull, and looking right from Mt Gabriel we have the bright glow of Ballydehob, Whiddy Island behind the hills (the floodlit white of a major oil terminal, Ireland's strategic oil reserve), then Bantry an extended yellow glare behind the hills and finally at the right edge (by now looking due North) and this side of the hills, the village of Church Cross.;

Looking up, above Mt Gabriel the bright smudge is the Andromeda Galaxy which, at 2.5 million light years away, is the most distant thing visible to the naked eye [see footnote]. Together with the Milky Way, it was chiefly what I wanted to catch in the photo, if possible. This is the “boring side” of the Milky Way which unfortunately is the only bit we get to see in the Northern Hemisphere winter. Even this, though, is evident to the naked eye from Ballylynchy. By the way, if you know your sky, you might also be able to spot the Triangulum Galaxy amongst all the noise. It's there.

Round 3a - 5/10 for composition, but much better execution: 8/10 I reckon

same scene with Ursa Major

This is the same scene as before, but at a different time of year (September, rather than the New Year of the Andromeda picture). The dominating night-sky feature is now not Andromeda or the Milky Way, but Ursa Major (The Plough) whose "saucepan end" is nearly vertical, pointing up towards Polaris, i.e. approximately Due North. Exposure: 45 seconds 24mm f/1.4 ISO 1600 with an AstroTrac.

It's not as pleasing a picture as the Andromeda/Milky Way one, but it is much better executed. The graininess is gone and the sky has a liquid quality that's beautiful. Compositionally it's markedly inferior, way too much deep black foreground, not enough sky. But, the sky colour is correct: orangeish at the horizon (light pollution) and above the horizon a greenish glow that's 558nm Oxygen-emission "airglow", if our eyes were sensitive enough they'd notice it. And finally the "fading to black" of the lovely West Cork night sky and its amazing starscape.

In due course, hopefully around Christmas 2017, I'll re-take this with the Milky Way and Andromeda (and Triangulum) back in. Hopefully my latest equipment addition, a heavy-duty Ball Head (Sirui k-40x) for the top part of my AstroTrac set-up, will allow me more easily to point the camera at the scene I want with fewer contortions - the orientation of the earth's equatorial plane this far North means the whole tripod-top set-up must be angled over at 39 degrees!

Round 3b – Round 1 re-done using lessons learned

colourful MW

My first Milky Way picture, the “horrid mess”, also deserved a re-run, essentially a practise–run to learn lessons. Again I used my Canon EOS 6D, my new 24mm Samyang and the AstroTrac: 51 seconds, 24mm at f/1.4 and ISO 1600. I also post-processed it better. As a result there’s lots of colour and little noise.

The bright star almost centre-frame is Altair, flanked by its two neighbours Tarazed (on the right) and Alshain. Looking down and slightly right, across the Milky Way cloud lane, notice a pair of little gems: lighter patches in the orange: open clusters NGC6633 and Graff’s Cluster, aka Tweedledum and Tweedledee. Open a larger version by clicking on the image and take a closer look.

Summary of Lessons Learned and applied so far: Round 1 to Round 3

I received some good “press” for my Round 2 pic and was pleased but I realised that, although not bad for a pure novice second attempt, it was still not very good. I determined to re-take both these pictures, and to get them right. Specifically, I needed to improve and expand my equipment (slightly), my shooting and exposure technique (quite a lot), my subject knowledge (somewhat) and my post-processing (a lot).

Equipment

My Canon EOS 6D is, apparently, a superb camera for this sort of work: full-frame sensor to make full use of wide-angle lenses; very low noise from “on-sensor noise suppression”, eliminating the need to take and subtract bias or dark frames (don’t ask).

The lens I used for Rounds 1 and 2, the Canon EF 24-70mm f/2.8L, is less ideal. True, it's an “L” lens (Canon’s pro line). But it’s a zoom, and zooms seriously distort stars. Also, it’s only f/2.8, which is barely fast enough. Even Canon’s and Nikon’s fast primes (non-zooms) are terrible for star distortion, and very expensive.

I got a Samyang/Rokinon 24mm f/1.4. Renowned for low distortion, it renders stars well and is only “quite” expensive. It’s manual-focus, but auto-focus doesn’t work on the night sky anyway. It will let in 4 times as much light as an f/2.8, so I could've achieved the same quality as the noisy Round 1 in only 7.5 seconds rather than 30, and without trailing. To eliminate the star-trails altogether, I got an AstroTrac, which is a beautiful mechanism that sits between two heads on a tripod. It rotates the camera to exactly counteract the movement of the stars around the sky as the Earth itself rotates.

Exposure and Shooting Technique

This is a huge subject, far too big to go into in detail here even if I understood it properly, which I don't yet (there's no end) so I'll simply suggest you visit the best-by-far and comprehensive description of the topic by Richard N Clark aka clarkvision.com . Armed with this new knowledge for Rounds 3a and 3b, 45-50 seconds (tracked) at f/1.4 and ISO 1600, produced nice results. Compare that to Round 2 for example, 30s at f/2.8 ISO 1600, which collected only 1/6-1/7th the amount of light.

The pair of image-extracts below from Round 2 and Round 3, of the same object, Mount Gabriel, demonstrates the difference that 6x the light makes:

......noisy Mt Gabriel.......... smooth Mt Gabriel

Notice the right-hand image now suffers from “streaking ground” rather than “streaking stars”, because of the sky-tracking. Depending on how important you think that is, you can either ignore it, or take a separate untracked frame and do some image-editing to overlay the untracked - and therefore sharp - bottom portion. You may think that’s “cheating”, but it's not: more about this a bit later.

Focussing deserves a paragraph of its own. Auto-focus at night doesn’t work, and it’s a bit hit-or-miss even in daylight. Focus needs to be absolutely spot on for stars. Set the camera to high sensitivity such as ISO 6400, find a suitably bright star and set “Live View” to 10x magnification. Focus by hand until sharp. Richard Clark advises going even further and using a magnifying glass as well (don't forget to change the settings back after focussing!).

Furthermore: shoot RAW, use a remote shutter release, set mirror lock-up and 2 second shutter delay to minimize vibration; long-exposure noise reduction should be OFF (not necessary for modern sensors) and the top dial set to "B" for “Bulb” (otherwise you're limited to 30 seconds); I’ve saved all these settings in a “Custom Setting” on the camera’s top dial.

Post-Processing Most people, using a modern digital camera, press the button, and Hey Presto! there’s a jpeg.

But the camera is doing a lot of work between recording a raw data file from the sensor and producing a viewable image file. The camera is making many decisions in the conversion process, arbitrarily deciding tone, contrast, saturation, black point, sharpness etc etc. Night-sky images are extremely dependent on these settings, and control needs to be taken over them from the very start. For example, "light pollution" will need to be "subtracted" from the image in almost all cases and at the outset of processing.

Shooting RAW makes the (albeit large) raw data file the camera’s primary output, preserving every scrap of information (every photon). Downloading the files into a “developer program” such as Adobe Raw Converter, which comes with Photoshop, or Adobe DNG Converter, which is free, allows me to manipulate the settings for conversion into an initial viewable image, and once downloaded (“developed”) further processing can be done in the application of choice, which in these cases was Photoshop. Learning these settings is a continuing matter of practise, research and trial-and-error.

Meanwhile, for now, I hope this has been interesting and perhaps of some help to any aspiring night-scene photographers.

Note about how astronomy pictures are produced

I alluded above to the fact that many people regard making images from anything other than single-shot frames as "cheating". However every single picture you see from NASA or the Hubble Space Telescope will be a stack of sometimes hundreds of "sub-images" overlaid on top of each other and signal-processed to extremes to extract the very few photons that arrive to the camera-sensor from the dimmest of subjects [present company excluded ;-) ], and to minimize the noise. If there's a foreground involved, such as a Milky-Way nightscape, it will have been taken "untracked" and pasted over the top in something like Photoshop to allow both stars and ground to appear "unstreaked" in a 30-60++ second exposure. There's simply no other way to do it.

**Some will tell you that actually the Triangulum Galaxy (aka Messier 33) slightly further away than Andromeda (if a mere 200,000 light years further can be called “slight”!!!!), holds that accolade. Look up Bortle Scale, which is a popular way of categorizing how dark a site is, to see that it uses the “naked-eye-osity” of Triangulum as a darkness-diagnostic. But this has been challenged lately, and I for one would need truly exceptional conditions to barely observe Triangulum with the naked eye. If, using the Bortle Scale, one inserts Andromeda (M31) instead of Triangulum (M33), it makes much more sense, for my mid-50s-year-old eyes at least. back up*

\$\endgroup\$
3
  • 2
    \$\begingroup\$ Hi Magnus, welcome to Photo.SE. Could you include some key details from your blog post in the answer? As it is now, if the blog post ever disappears, this no longer answers the question. It's fine to use external links for additional information and reinforcement, but answers should still provide an answer without the external links. \$\endgroup\$
    – AJ Henderson
    Commented Nov 10, 2017 at 17:39
  • \$\begingroup\$ Thanks for that, bit of a newbie on here, I will post an extract from my blog... \$\endgroup\$
    – Magnus
    Commented Nov 15, 2017 at 11:45
  • \$\begingroup\$ OK I've transcribed my blog post (blog was only completed a few days ago anyway so it's essentially new material). Apologies for some obvious "link phrases" that don't actually link, I'm only allowed a few links being a brand-new forum member. \$\endgroup\$
    – Magnus
    Commented Nov 15, 2017 at 14:00
1
\$\begingroup\$

Long exposures, fast lenses, low background light, clear skies and having the milkyway overhead at the time. This typically takes very long exposures. In this photo, it was likely a 30 second exposure with a flash used to expose the couple on the bench quickly.

If there is too much background light from nearby cities or the moon, then it won't be possible as it will be washed out. Similarly, if the milky way isn't overhead at the time, you will get a mostly black star field.

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

Well lets start with post processing as you specifically mention it.

You can bump up saturation or vibrance (use the later if available). Here's one I did in GIMP.

GIMP processed version of OP photo

The method I used was to :

  • load the image
  • create a duplicate layer
  • add a layer mask based on the inverse greyscale (this is available from the layer mask dialog). This will "protect" the image from becoming swamped by saturation increases.
  • Select the image in the layers window (when you create a layer mask GIMP leaves the mask as active, and you need to reselect the image itself to make changes to it).
  • Open the "Hue Color" dialog from the "Color" menu and simply ramp up saturation all the way.
  • Do a "merge down" operation (from the layer mask menu).
  • To get this image I did this again to get another "bump" is appearance

Now the result does look a little better but it's not a great section of the milky way to start from.

Other people have commented on technique and I'd just add these remarks :

  • Your camera is fine - honestly, people have been doing this since the beginning of consumer ASPC DSLRs and they get fine results. It does require practice
  • Noise. I'd suggest you read this page about image types and how to use them in Astronomy photos. Note in particular the marks about dark frame subtraction.
  • Free image processing software designed for Astronomy : try IRIS.
\$\endgroup\$

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.