Jupiter’s ice-covered ocean moon Europa floats above the planet’s Great Red Spot in this 1979 image from Voyager 1. This is a mosaic of several images in orange and violet filters. The scene is about 22,000 miles across.
Voyager 1 ISS NAC, 0.96 sec (orange), 0.48 sec (violet), f8.5
Image: NASA/JPL/Alexis Tranchandon
I’m both an amateur and professional photographer. As an amateur, my photography is a source of great enjoyment, but nothing special. But my professional photographic work is more interesting. As a member of several large teams of planetary scientists and engineers, I get to photograph other worlds, from up close.
The cameras we use cost millions of dollars and are attached to some of NASA’s robotic interplanetary spacecraft, which take them to amazing places. But photography is photography, and the fundamentals of how we capture the interaction between light and matter to generate images are the same. I’m surprised how often the two kinds of photography overlap.
"But photography is photography."
We explore the worlds of our solar system using spacecraft bristling with many tools – magnetometers, mass spectrometers, radar, and so on. But cameras provide the most accessible information on the worlds we explore and are incredibly versatile in the range of phenomena they can capture. It’s rare for a spacecraft to head out to the planets without cameras of some kind.
Planetary Photoshoots
Our first visit to a new world is usually a flyby, using a spacecraft that doesn’t even slow down as it passes its target but grabs what images and other data it can in the precious few minutes or hours when it’s close by.
Every detail is planned months or years beforehand and rigorously tested before being uplinked to the spacecraft for execution. The flybys themselves are then a matter of watching and waiting, hoping everything goes smoothly, and excitedly pouncing on the new data when it hits the ground.
Some of the science team for the Lucy asteroid mission, at the moment we got our first look at images from the Lucy cameras that showed a moon, unknown until then, orbiting the asteroid Dinkinesh. The gobsmacked author is seated, center, in the blue t-shirt.
Image: Stuart J. Robbins.
That first picture of Dinkinesh and its moon Selam (left), from Lucy’s low-resolution tracking camera, is responsible for the reaction above. Much better pictures (right), from our long-lens LORRI camera were downlinked a couple of hours later. Dinkinesh is about 0.4 miles across.
Left: Lucy TTCAM, 1/1250th sec, f2.95
Right: Lucy LORRI, 1/500thsec, f12.6
Image: NASA/Goddard/SwRI/Johns Hopkins APL/ASU/NOIR
Later, if we can, we return with spacecraft that carry the big rocket engines and fuel needed to get into orbit for an extended stay, often spending years in detailed exploration or even landing and roving for a much closer look.
Cameras
NASA was, by necessity, an early adopter of digital camera technology. The first close-up pictures of Mars, taken by Mariner 4 in 1965, were obtained with an analog vidicon camera, but were digitized (200 x 200 pixels, 6-bit) for transmission back to Earth at a blistering 8 bits per second.
This technology (upgraded to 800 x 800 pixels and 8 bits) was used until the late 1970s, and the Voyager mission’s astonishing images of the outer gas giant planets and their moons, and that final, famous, “pale blue dot” look-back image of the Earth, were all digitized vidicon images.
The first interplanetary photograph, of the Martian horizon, was taken on July 14th, 1965. The actual image is on the right. The scene is about 300 miles across. On the left is the first rendering of the image, made by an impatient engineer by hand-coloring pasted strips of printouts of the data numbers.
Mariner 4 TV camera, 1/5th sec, f8
Image: NASA/JPL-Caltech/Dan Goods
But starting in the 1980s, long before they were adapted to consumer cameras, solid-state CCD detectors became the norm, bringing greatly increased sensitivity and image quality. Camera designs tend to be conservative, though, because reliability is an overriding concern when the nearest repair facility is a billion miles away, and the tried and true often beats the innovative.
CMOS detectors, originally developed for NASA use, are now becoming common, but we chose a 1 Megapixel CCD detector for the LORRI telephoto camera included on our asteroid mission, Lucy, which launched in 2021. The format may be small and the technology old-fashioned, but we’d flown the LORRI camera before, on the New Horizons mission to Pluto, and we knew that it would work.
Pluto, photographed in approximate natural color (left) and enhanced color that includes near-infrared data (right) by the New Horizons spacecraft on July 14th, 2015 (coincidentally, 50 years to the day after that first Mars image). Pluto’s diameter is 1,470 miles.
New Horizons MVIC, 0.60 sec, f8.7
Image: NASA/Johns Hopkins University Applied Physics Laboratory/Southwest Research Institute/Alex Parker
For larger format images, we often use long, skinny, “pushbroom” arrays. A scan mirror, or the entire spacecraft, moves to sweep the array across the target while the array is read out in lockstep with the scene motion. Pushbroom has the advantage over large framing arrays in that the detector is much more compact, and it’s also easy to incorporate color by using a series of parallel arrays, each with its own color filter, which scan over the target in turn.
Because maximizing detail is paramount, focal lengths tend to be long, and most planetary cameras are really telescopes. MVIC is the “wide angle” camera on New Horizons, but has a field of view of just 5.7 degrees (350 mm equivalent focal length), while its narrow-angle traveling companion, LORRI, has a 0.29-degree field (7000 mm equiv. focal length). Focal lengths are limited both by camera weight and by how steadily the spacecraft can track the target for the necessary exposure times.
Color
The simplest planetary cameras, like the New Horizons and Lucy LORRI cameras, are monochrome. For color, scanning with a set of linear pushbroom arrays, each with a different-colored overlying filter, is often used, as mentioned above. Other cameras obtain color images with Bayer-type filter arrays or filter wheels that step through the wavelengths in turn, as in the Voyager image of Jupiter and Europa above.
"When the nearest repair facility is a billion miles away, the tried and true often beats the innovative."
Matching human color vision is usually less of a priority than choosing the most scientifically diagnostic wavelengths; the New Horizons MVIC camera carries blue and red filters, but not green, and has two near-infrared filters, one tuned to a wavelength (0.89 microns) that is strongly absorbed by the frozen methane that’s abundant on Pluto’s surface.
Reconstructing “natural color” images from the resulting data can thus be tricky and somewhat subjective. But the aim is generally to show real variations in the color of the scene, whether or not they correspond precisely to what the eye would see, as well as to produce something aesthetically pleasing. Including wavelengths beyond human vision increases color contrasts and reveals patterns that the eye would miss, as with the Pluto example above.
Sometimes, color saturation is cranked up to bring out subtle features. We try to label released images to make these distinctions clear, though the provisos often get lost when the images are reproduced.
Color image releases of Jupiter’s volcanic moon Io, 2250 miles in diameter, from Voyager in 1979 (left), Galileo in 1997 (middle) and Juno in 2023 (right), illustrate the vagaries of representing planetary colors. Colors vary due to the different sets of color filters used and the preferences of the image processors. Voyager, in particular, missed the red color of the huge oval ring of volcanic fallout surrounding the Pele volcano (lower-right center, left image and lower-left, center image), because Voyager’s vidicon detector was blind to red light. Spot-the-difference fans can enjoy finding the changes wrought by volcanic activity between the Voyager image and the Galileo image, which cover much of the same terrain.
Left: Voyager ISS NAC, 0.49 sec (orange), 0.36 sec (blue and violet), f8.5 (NASA/JPL)
Middle: Galileo SSI, 0.4 sec (violet), 1/20th sec (green), 0.26 sec (infrared), f8.5 (NASA/JPL/University of Arizona)
Right: Juno Junocam, red, green, blue filters, f3.2 (NASA/JPL-Caltech/SwRI/MSSS/Gerald Eichstädt)
Lighting
All interplanetary photography is outdoor photography, and the sun is our primary light source. We have one variable that earth-bound photographers don’t need to worry about, which is the varying brightness of the sun depending on our distance from it.
"All interplanetary photography is outdoor photography."
Pluto, which the New Horizons spacecraft flew past in 2015, was then 33 times further from the Sun than the Earth is, and sunlight was 1000 times weaker, something like the illumination in a cozy terrestrial restaurant. This effect is predictable, and our camera focal ratios and exposure times are designed to handle it.
New Horizons’ MVIC camera had no trouble obtaining those color images of Pluto in that dim light. But in 2007, when New Horizons flew past Jupiter en route to Pluto, and we were six times closer to the sun, and the sun was forty times brighter, MVIC was hopelessly overexposed, and we didn’t get any decent color pictures of Jupiter.
Part of the bizarre surface of Europa, seen by the Galileo Jupiter orbiter, under high sun (left) and looking very different in a closer image mosaic of the region in the yellow square, taken with oblique illumination (right). The yellow square is about 220 miles across. The left-hand image shows compression artifacts, a consequence of Galileo’s broken main antenna, which required severely compressing images for downlink at very low rates through its backup antenna.
Galileo SSI. Left: 1/240th sec. f8.5; Right: 1/80th sec f8.5
Image: NASA/JPL-Caltech
With the sun as our primary light source, we don’t have the luxury of repositioning it to our liking. But like any landscape photographer, we can control the lighting by choosing the timing and viewpoint for our images.
As on Earth, long shadows provide the most dramatic landscapes, highlighting subtleties in topography that would be washed out with the sun overhead. So low sun is best for understanding the lie of the land, while high sun is best for capturing brightness and color variations that give clues to what the surface is made of.
The night side of Saturn’s distant moon Iapetus (912-mile diameter) photographed in Saturn-light. The spacecraft rotated to track Iapetus during the exposure, streaking out the images of background stars.
Cassini ISS NAC, 82 sec, f10.5
Image: NASA/JPL/Space Science Institute
Space being black, our sources of indirect light are limited, and space lighting tends to be direct and harsh. But sometimes, we can use indirect lighting to see where direct sunlight can’t reach.
Sunlit topography can reflect light into nearby shadowed regions; a NASA camera called ShadowCam on the Korean KPLO lunar orbiter exploits this indirect light to look for signs of ice in frigid lunar polar crater bottoms that never see direct sunlight.
Other nearby worlds can also provide indirect illumination, just as Earthshine illuminates the dark side of the crescent moon. Our best images of some parts of Saturn’s moon Iapetus were obtained, with very long exposures, using Saturn-shine. Saturn’s enormous rings provide dramatic indirect lighting on Saturn’s night side, which I’ve always found particularly beautiful, providing a soft light rarely seen in space scenes.
Ring-shine illuminates the night side of Saturn, as seen from Voyager 1 in 1980. Saturn’s shadow cuts across the rings on the left. The scene is about 50,000 miles across.
Voyager 1 ISS WAC, 15.4 sec, f3.5
Image: NASA/JPL
Saturn-shine illuminates the night side of Saturn’s active moon Enceladus. Jets and curtains of ice particles, erupted from geyser-like fissures in Enceladus’ south pole, rise up out of Enceladus’ shadow to catch the direct sunlight. The scene is about 250 miles across.
Cassini ISS NAC, 3.2 sec (red), 3.8 sec (green), 18 sec (blue), f10.5
Image: NASA/JPL/Space Science Institute/Gordan Ugarkovic
Composition
We’re rarely doing this for art’s sake; our goals are utilitarian, pursuing the best possible combination of detail and coverage to understand our targets. Considering where we are and what we’re looking at, though, the results are often stunning. And we still make aesthetic choices when choosing which images, or parts of images, to highlight for early public release.
Creative cropping. A parting shot of Pluto from New Horizons (left), and the most spectacular part of the image (right), which we chose for early public release. We would have loved to take this picture in color, but couldn’t spare the time to store the additional color data. Pluto’s diameter is 1470 miles, and the enlargement on the right is 230 miles across.
New Horizons MVIC, 0.40 sec, f8.7
Image: NASA/Johns Hopkins University Applied Physics Laboratory/Southwest Research Institute
Saturn’s moon Dione, behind Saturn’s edge-on rings, seen below the south pole of the moon Rhea (top). Despite appearances, this is a real single-exposure image, not a composite. Diameters of Rhea and Dione are 950 and 700 miles, respectively.
Cassini ISS NAC, 1/12th sec, f10.5
Image: NASA/JPL/Space Science Institute
Occasionally, when the timeline is relaxed and we have resources to spare, we have the luxury of planning photos primarily for their aesthetic appeal. The Cassini Saturn orbiter took many images during its 13 years in Saturn orbit that were designed primarily to capture dramatic alignments of Saturn, its moons, or rings.
When New Horizons flew past Jupiter in 2007, we enlisted the help of amateur space enthusiasts to help us compose some of these scenic shots, including this alignment of the ice-covered ocean moon Europa with Io, its volcanic sibling.
Jovian moons Europa (left) and Io (right), imaged by New Horizons shortly after flying by Jupiter on the way to Pluto. The night side of Io is illuminated by Jupiter. Three volcanic eruptions, one with the red glow of incandescent lava at its center, can be seen on Io. This image is a composite of a high-resolution monochrome image from the LORRI camera with color from the lower-resolution MVIC camera. The diameters of Europa and Io are 1940 and 2260 miles, respectively.
New Horizons LORRI, 1/12th sec, f12.6; New Horizons MVIC, 0.6 sec, f8.7
Image: NASA/Johns Hopkins University Applied Physics Laboratory/Southwest Research Institute
Exposure
Like all photographers, we think a lot about exposures. Aperture is fixed by the camera system – all our subjects are at infinity, and we always shoot wide-open. For exposure times, the usual trades apply; we don’t want to blow our highlights, but we want to minimize noise and retain shadow detail.
We also want to minimize motion blur and camera shake, so the motion in the scene, and the steadiness of our platform, are important. The Cassini spacecraft that took the above long-exposure Iapetus image was spectacularly steady, other craft are less so.
"Like all photographers, we think a lot about exposures... we don’t want to blow our highlights, but we want to minimize noise and retain shadow detail."
But we have some unique challenges. First, we can’t make adjustments on the fly – shutter lag is a big deal when your camera is up to several light-hours away, and with flyby missions, our subjects would be long gone before we could tweak our exposures. And we rarely use auto exposure, mostly because of its unpredictability.
Second, we often don’t know how bright our targets are going to be if we’ve never seen them up close before. So we often fall back on the old photographer’s standby of exposure bracketing, or we increase dynamic range by taking a bunch of short exposures and stacking them later.
Exposure challenges. Left: The Kuiper Belt object Arrokoth, severely underexposed to limit smear, given the feeble sunlight at 42x the Earth’s distance from the sun. Noise was reduced in the final product (below) by combining nine of these individual images. Arrokoth is 22 miles long.
New Horizons LORRI, 1/40th sec. f12.6
Image: NASA/Johns Hopkins University Applied Physics Laboratory/Southwest Research Institute
Right: The asteroid Dinkinesh, with blown highlights in the center of the image. Exposure choice here was a deliberate gamble, as we had to choose a single exposure for a long sequence of images with varying lighting. We chose to risk overexposure when the sun was behind us, as in this image, in order to optimize exposures in the rest of the images (including the image of Dinkinesh and Selam above), where Dinkinesh was fainter. Arrokoth and Dinkinesh are made of stuff with similar intrinsic brightness, but Dinkinesh is 20x closer to the sun, and sunlight is 400x brighter. Dinkinesh is about 0.4 miles across.
Lucy LORRI, 1/500th sec. f12.6
Image: NASA/Goddard/SwRI/Johns Hopkins APL
The final processed closest Arrokoth image (left and right), flanking an earlier image (center) taken in a similar way from a different angle, from greater distance. The images are arranged so Arrokoth can be viewed in stereo, either via parallel viewing (left and center) or cross-eyed viewing (center and right).
Image: NASA/Johns Hopkins University Applied Physics Laboratory/Southwest Research Institute
Summary
The adventure of planetary exploration continues. I’ve focused on U.S. missions here, which currently face serious budget threats. But this is now an international effort, with players including Europe, Japan, China, India, and several smaller nations. In addition to a fleet of spacecraft currently exploring Mars, spacecraft are currently on their way to Mercury, several asteroids, and Jupiter’s moons Europa and Ganymede, with launches planned soon to the Martian moons, and Saturn’s moon Titan. All of them carry cameras, and hold the promise of amazing photographic opportunities to come.
Additional notes:
Spaceflight is plagued with acronyms, and I haven’t attempted to spell them out here. For explanations and much more detail, follow the included links for the various cameras. The raw image data are generally available for download from the NASA Planetary Data System- see, for example this excellent search tool for outer planet images. Extensive processing of the raw data, often done by enthusiastic amateurs, is used to create many of the images here, but they remain true to the original data. If you have any questions, feel free to ask me in the comments.
John Spencer is a planetary scientist at the Southwest Research Institute in Boulder, Colorado. A member of the science teams for NASA missions to Jupiter, Saturn, Pluto, and the Trojan asteroids, he has led the science planning for several planetary flybys. His earthbound photography includes documenting the excitement of planetary encounters, including the New Horizons encounter with Pluto.