On this page
See also the Low-Light Checklist
Last updated October 22, 2009
Its enduring beauty is reason enough to try, but the moon can be a challenging digital subject:
Blooming aside, film photographers grapple with much the same issues, but 35 mm film SLRs tend to be well-equipped for such challenges. Some digital cameras aren't, but with the proper resources in hand, important digital advantages will accrue—including the ability to check your work at the scene and direct access to digital post-processing. Less obvious a benefit is the LCD's overly bright image, which exaggerates lens flare and helps you center up the moon to eliminate it.
Once you've captured your quarry, the Sky & Telescope photographic moon map will help you identify its features. It's amazing how much lunar geography you can see with even the unaided eye—once you know what's there.
To come to terms with the moon on the digital side, you'll need at minimum a camera with
These features tend to be found only on higher-end consumer-grade digital cameras, but they're becoming more and more affordable.
Beyond the camera, you'll also need
If you persist in shooting the moon, you'll sooner or later yearn for more magnification. Teleconverters are certainly a start, but you'll need a spotting scope or a telescope to fill the frame with the moon. Such instruments can often be used to great advantage with relatively simple afocal camera mounts, as explained here and here. A visit to Michael Myers' digital moon gallery will give you an idea of what can be done with a 60x Swift Telemaster spotting scope and a 3MP Nikon Coolpix 990 via an afocal mount.
What You Won't Need
Short of telescopic magnifications, you generally won't need a motor-driven telescope mount to stop the moon in its tracks, as explained below.
With a maximum angular size of little more than 0.5°, the moon's a small target by any standard. Since getting closer isn't all that practical, magnification is the only way to make it bigger, but short of a telescope or spotting scope, you'll have a hard time getting a frame full of moon with a digital camera.
Why bother filling the frame with the moon?
Higher magnifications will require greater exposures for the same lunar image brightness, but that's usually a simple adjustment.
Calculating Your Frame-Filling Focal Length
So, just what kind of focal length would it take to fill your camera's frame with the full moon? "A very long one" is the short answer. The long answer follows.
According to a table in Dave Martin's article Astrophotograhy for Beginners, the full moon's image turns out to measure 0.91 mm at the image receiver for every 100 mm of absolute focal length applied. The ratio holds for digital and 35 mm film cameras alike, but for digital cameras, you'll need to use the actual focal length, not the 35 mm equivalent.
If you know the short dimension of your CCD's active sensel array, you can use the ratio 0.91/100 to estimate your frame-filling focal length for the full moon, as illustrated below. If your CCD has a 4:3 aspect ratio, and most do, the short dimension is 60% of the effective diagonal, which just happens to complete a 3:4:5 triangle.
Andrzej Wrotniak's excellent moon photography article offers an even simpler approach, which I've taken the liberty of modifying ever so slightly for computational ease: To approximate the percentage of the short side of your frame filled by the moon at any given 35 mm equivalent focal length, simply divide that focal length by 30 mm. Thus at 300 mm, the moon will fill ~10% of the frame. In other words...
The active surface of my Oly C-2020Z's 2.1MP, 4:3 CCD measures 4.8x6.4 mm with an 8.0 mm diagonal. To fill its frame with the full moon, I'd need a whopping final actual focal length of 100 x (4.8 / 0.91) = 527 mm! On a 35 mm camera, that focal length is equivalent to 2625 mm, an impossibly tall order for any camera lens. With an actual focal length of only 19.5 mm at full zoom, my C-2020Z would need a 27x increase in focal length to fill its frame. That kind of magnification falls squarely in telescope or at least spotting scope territory.
That kind of math led Mike Myers to team his 3MP Nikon CoolPix 990 up with a 60x Swift Telemaster spotting scope to capture the spectacular moons you'll find in his impressive digital moon gallery.
It may take a telescope to fill the frame on most digital cameras, but worthwhile moons can still be had with one or more teleconverters:
Of course, many other camera-teleconverter options are available.
Acknowledgement: Thanks to Jim Martin of RPD for the frame-filling concept and for a link to Dave Martin's table establishing the 9.1/100 "moon image size to focal length" ratio.
As always, motion blurring happens on both sides of the camera. A sturdy tripod and shake-free triggering—e.g., with an IR remote control—will eliminate camera shake, but the moon's never going to hold still. The ultimate defense against lunar motion is an equatorial camera mount with a rotation-canceling right-ascension motor drive, but drive mounts are usually necessary only you're shooting through a telescope. If the moon is your only subject, a shutter speed fast enough to freeze the moon is usually very easy to arrange, but if other image elements must be accommodated with the same exposure, you may have a problem on your hands.
The question is, how fast is fast enough?
The Moon's Apparent Motion
To observers rotating with the Earth, the moon appears to move along the Ecliptic at ~15 arc seconds per second. That amounts to 1/120th a moon diameter per second, or a full moon width (0.5°) every 2 minutes. The apparent direction of the moon's motion depends on which hemisphere you're in.
The moon's apparent motion is real enough to smear the its shape and blur its surface features when telescopes or long exposures (mandated, perhaps, by other image elements) enter the picture. But for routine moon work at reasonable shutter speeds, lunar motion will seldom be a serious hurdle.
Suppose you were to fill the 1600x1200 frame of your digital camera with the moon using a very long (527 mm) lens system. With 1200 pixels now spanning its diameter, the moon will move exactly 1 pixel during a 1/10 sec exposure.
That's a good benchmark to work from. Adding more magnification (more pixels per moon diameter) or further prolonging the exposure would certainly result in a smear. In all likelihood, however, you'll likely be working with substantially less magnification and much shorter exposures.
As you can see, at 1/100 sec and faster shutter speeds, the moon's motion should be undetectable at all but telescopic magnifications. With a reasonably fast lens, manual exposure, and the ever-pressing need to avoid blooming, slower shutter speeds should come into play only for thin crescent moons.
The unsharpened sample moon images below support this approach. Each moon there is roughly 80 pixels across.
For the moon, manual exposure, heavy bracketing and checking your work at the scene are your best bets on the exposure front, but where to start?
First, for maximum sharpness of lunar details, settle on an aperture near the resolving power sweet-spot for your lens. For many lenses, that's 2 full stops down from wide open at the chosen focal length.
Plan on controlling exposure with shutter speed based on the sunny f/16 rule for the full moon, but be prepared to depart from that for other lunar phases and settings. Reaching shutter speeds fast enough to stop the moon's motion shouldn't be a problem given the general need to underexpose to preserve surface detail. To minimize noise, select the smallest feasible ISO.
If you can't reach the proper exposure at your optimum aperture, consider a neutral density (ND) filter in lieu of stopping down the aperture, but watch out for flare. (Since moonlight is largely unpolarized, a polarizer will function nicely as a 1- to 2-stop ND filter here.)
The lens on my C-2020Z is sharpest at around f/5.6 at full zoom. Shutter speed bottoms out at 1/800 sec on this camera, but luckily, f/5.6@1/800 gets me where I need to go with any moon.
Unless you've managed to fill most of the frame with the moon, there's little point in metering it with your camera, even with spot metering at full zoom. At 0.5° across, the full moon swims in the mostly black 3° solid angle my fairly typical Oly C-2020Z spot meter covers at full zoom—even with a 1.7X teleconverter mounted.
Don't Bloom That Moon!
To avoid blooming an auto-exposed moon, you'd need to apply heavy negative exposure compensation (EC), but the ±2 stops of EC leeway available on most cameras won't get you to a detail-preserving moon exposure in most instances. The onset of blooming is 3 stops above my favorite full moon exposure (f/5.6 @ 1/650), and the typical auto-exposure is at least that many stops higher still when the moon occupies only a small part of the frame.
Manual exposure based on the sunny f/16 rule minus part of a stop (to taste) is the most reliable way to shoot the full moon at sub-telescopic magnifications, as the next section illustrates. Lesser phases require more exposure in a non-linear way, as shown in the phase table below.
For a full moon high in a clear, dark sky, the sunny f/16 rule is a good starting point for exposure but will strike many as too dark at first glance. The unaided brain-eye perceives a substantially brighter moon, but this rule avoids blooming and nicely preserves lunar surface detail. Since the moon's surface is generally composed of materials darker than the medium-toned subject the sunny f/16 rule assumes, it's not unreasonable to bump the exposure 0.3-1.0 stops from there.
Lower exposures also help keep lens flare at bay when the shot demands an off-center moon.
Keep in mind that increasing magnification spreads the light of the moon over a larger area on your image receiver. Each CCD sensel will receive fewer photons as a result. If you work out a bare-camera exposure at full zoom and then slap on a teleconverter, you'll likely have to increase exposure to maintain the same image brightness.
Fill the frame enough, however, and spot metering will begin to give you workable exposures.
Pick a Moon, Any Moon
The moon gallery below shows a reasonable range of sample manual exposures at ISO 100 and f/5.6, the optimum aperture for my lens. I captured this particular moon 31 hours before full and 24 hours before perigee on 2/6/01, but the exposures are probably within 0.1 EV of the results I would have obtained with an average full moon, as the phase table below suggests.
If you find an exposure you like here, use reciprocity or an EV table as needed to determine an equivalent optimized full moon exposure for your camera. If you settle on a different ISO, you'll have to adjust your exposure accordingly. (For example, if you choose ISO 50, open the aperture or increase the shutter speed by a stop, or subtract 1 from the chosen EV before consulting the EV table.) When in doubt, underexpose, and don't forget to bracket!
Technical Notes: Oly C-2020Z at full zoom with a B-300 1.7x teleconverter (180 mm final focal length), ISO 100, aperture f/5.6, 4:1 (SHQ) JPEG recording with in-camera sharpening disabled, tripod with IR remote triggering, no post-processing other than cut-and-paste, saved here full size in PNG format with lossless compression. The lunar north pole is to the left; Mare Crisium is the small round dark spot near the limb at 12:30. Compare these images with the Sky & Telescope photographic moon map.
I can't see motion in any of these images, even in the 1/100 sec sample.
The f/5.6 @ 1/650, EV 14.3 exposure is my personal favorite, largely because I like the detail it preserves. It also happens to be the true sunny f/16 exposure for my camera. (Minimum ISO is nominally 100 on the C-2020Z, but 80 is more realistic on mine.) On close inspection, surface detail starts to dissolve at around f/5.6 @ 1/400, at least to my eye. This loss of detail is not a manifestation of motion blurring but of blooming.
When Bill Paire first told me via RPD that sunny f/16 exposures work for the full moon, I was as puzzled as I was pleased to find that an already familiar rule applied. Earthly sunlit subjects are typically at least 10,000 times closer than the moon, which orbits over 380,000 km away. Doesn't that difference in distance count somehow?
Understanding what doesn't affect exposure can be as handy as understanding what does.
For similar reasons, earthshine falling on a new moon is about as bright as moonlight falling on the earth. With this rough equivalence comes an opportunity to capture pre-dawn landscapes with a earthlit new moon in a single exposure, as detailed below.
The full moon is an imposing sight, to be sure, but other lunar phases have their charms. To portray craters and terminator features in maximum relief, you can't beat a side-lit quarter moon. (Surface detail is suppressed when the moon is full, as it is with any front-lit subject.) And for reasons known only to my limbic system, thin crescent moons bathed in earthshine move me the most.
Note: As seen from the northern hemisphere, the sun respectively illuminates the right (south) and left (north) sides of the moon in the 1st and 3rd quarters. In the southern hemisphere, the opposite obtains.
This section leverages the full moon manual exposure worked out above by developing relative exposures for other phases of the moon. For a variety of reasons, total lunar brightness falls off quite rapidly before and after the full moon—faster than one might expect from the fraction of the lunar disc illuminated—so it's only a matter of how much more exposure to throw at phases other than full.
As we'll see, a quarter moon needs 2-3 stops more exposure and a medium crescent up to 7 stops more relative to a properly exposed full moon.
A First-Order Lunar Brightness Model
Planetary brightness is proportional to 10-magnitude and therefore on first order to F5/2, where
This brightness model can be used to estimate exposure corrections for all lunar phases relative to a proper full moon exposure. So,
I calculated these starting-point exposure corrections for lunar phases high in a dark sky. Once again, err on the side of underexposure, and don't forget to bracket!
* Technical Notes: The "Days Out" column gives the number of 24-hour days before or after the full moon, which may or may not occur at local nighttime. The mean lunation is 29.53 days.
These calculated corrections jibe reasonably well with those offered in Michael Covington's lunar exposure tables in Astrophotography for Amateurs and in Astronomy Magazine's Photographing the Sky article.
The moon is best observed and most clearly and reliably photographed high in a dark sky. All the lunar exposures worked out thus far assume such conditions, but "ideal" conditions aren't always available. Nor are they routinely desirable from a photographic point of view, as this article's featured photograph attests.
Different lunar phases reach maximum altitude above the local horizon at different times of the year, as shown for the northern hemisphere in the table above. For the full moon, this occurs at the winter solstice, but the 1st and 3rd quarters are highest at the vernal and autumnal equinoxes, respectively. Conversely, the new moon attains maximum altitude at the summer solstice while waxing and waning crescent moons 3-4 days out from the new moon are highest ~6 weeks before and after summer solstice, respectively.
Full or otherwise, "off-peak" moons captured below maximum altitude need more exposure than shown in the table above, particularly those near the horizon. How much more is hard to say. Photographer Tony Sparado writing via RPD finds that his near-peak February full moons come out best with a sunny f/16 exposure, but his best August full moons, much lower in the sky, require 3 more stops.
I don't have enough data points to propose an altitude-aware exposure rule.
The Moon Illusion
While on the subject of lunar altitude, let's take a moment to dispel the moon illusion—the larger apparent size of the moon when it's low in the sky. No one denies that the perceived size of the moon decreases with altitude, but the measured angular size of the moon does not. Why the illusion occurs is still under debate, but contrary to popular belief,
Under certain conditions, atmospheric distortion can narrow the apparent vertical size of the moon near the horizon, but the horizontal size remains constant at around 0.5°. The result is a moon that may look a bit fatter horizontally but is actually smaller in overall apparent size.
A moon filling only a small portion of the frame seldom stands alone as a photographic subject. Adding other interests to the scene will generally make for a better photograph when you're working with less than telescopic magnification.
Unfortunately, photographing the moon in company tends be complicated tremendously by the need to accommodate competing scene elements, surrounding skylight, and other atmospheric effects related to low altitude, including scattering.
At certain times of year, you can capture spectacular moons against largely sunlit foregrounds with a single compromise exposure. Ansel Adams did just that in his famous "Moonrise, Hernandez" above. (Note that Adams darkened the sky with a dark red filter in this B&W photograph. You could do the same or use channel mixing in post-processing instead.) Prepare to struggle with adequate foreground exposure while maintaining a shutter speed fast enough to freeze the moon to keep it sharp and round. With careful spot metering of the moon and foreground, generous bracketing and frequent result checks at the scene, you may well find a workable compromise exposure, but you won't have much time to do it.
Between moonrise and dawn several days before the new moon, you have a fleeting opportunity to photograph pre-dawn foregrounds with a moon illuminated by earthshine rather than direct sunlight. The moon and foreground exposures will be similar then because the intensity of earthshine falling on the moon roughly matches that of moonlight falling on the earth. This is best done around the end of April, when the thin waning crescent moon reaches its maximum altitude.
Most of the time, however, you'll be working with a sunlit moon and a largely moonlit foreground. If flash can't adequately illuminate the foreground under these circumstances, you'll need 2 separate tripod-mounted exposures in quick succession to do justice to both—one near sunny f/16 for the moon and another at vastly different low-light settings for the foreground. You'll need a moon-stopping shutter speed only for the moon exposure since you'll be masking out the moon in the foreground shot.
Combining moon and foreground exposures in post-processing using layers and suitable masking techniques will give you a crisp, well-exposed moon and a well-exposed moonlit foreground.
Graduated Neutral Density Filters
You might be able to reduce the disparity between moon and foreground exposures with a graduated neutral density (GND) filter, but the slow shutter speeds typically dictated by moonlit foregrounds will usually be too slow to freeze the moon.
The moon looks simply marvelous in blue, but bright skylight can easily drown out any daytime moon.
Unfortunately, a polarizer won't do much good around full or new moons, when the camera, moon, sun and the air scattering skylight around the moon are all roughly aligned. Darkening the blue channel in post-processing might help moon-sky contrast, but I haven't tried it.
Acknowledgement: Thanks to Tony Sparado of RPD and Dark Alley for some data points and tips on shooting off-peak moons.
(See also the home page links.)
Astrophotography for Amateurs—Michael Covington's deservedly best-selling and widely-acclaimed book covering all aspects of astrophotography. In fact, I've learned a lot about photography and post-processing in general here. (2nd edition, Cambridge University Press, 1999)
Astrophotograhy for Beginners—an excellent introduction by Dave Martin of the Colorado Springs Astronomical Society
Astrophotography with Digital Cameras and Camcorders—an introductory article by Astrophotography for Amateurs author Michael Covington.
Complete Sun and Moon Data—for any day you choose, a service of the U.S. Naval Observatory. A planning tool for moon phases, moonrises and moonsets. Another useful page gives dates and univeral times (UT) for the primary moon phases for any year between 1990 and 2005.
Norton, A.P., Norton's Star Atlas, Sky Publishing Corporation, Cambridge, MA, 1978.
Phases of the Moon—the U.S. Naval Observatory's beautifully illustrated exploration of the phases of the moon, including an animation with some surprises.
Photographing the Sky—a concise online Astronomy Magazine guide on sky photography with a 35 mm camera readily transferable to the digital side.
Sky & Telescope photographic moon map—a great help in identifying lunar features.
Take a picture of the moon—physicist and digital photography enthusiast Andrzej Wrotniak offers some very practical tips on capturing the moon with any camera.
The Moon—Michael Myers' impressive lunar photography and information site featuring images made with a 60x spotting scope.
Unless explicitly attributed to another contributor, all content on this site © Jeremy McCreary