Posted on October 17, 2017
One of the greatest benefits digital photography has over film is the ability it provides to check an image’s exposure at capture (when you can do something about it). But as photographers, we rely so much on our eyes that it’s sometimes difficult to accept that they’re not always right. We take a picture, look at it on the LCD, and decide whether or not it’s perfect without even considering that there might be a better way. But where exposure is concerned, there is indeed a better way.
That’s because every picture you click, whether raw or jpeg (but especially raw), contains more shadow and highlight information than the jpeg that appears on your camera’s LCD preview screen can reveal. Fortunately, our digital cameras give us a tool that tells us about that missing exposure info: the histogram.
A histogram is a graph of the tones in an image. While I imagine that any graph has the potential to evoke high school science trauma flashbacks, a histogram is really quite simple, simple enough to be read and interpreted in the blink of an eye. And not only is your histogram easy to read, it’s really the most reliable source of exposure feedback.
When an image is captured on a digital sensor, your camera’s “brain” samples each photosite (the individual pixels in the megapixel number used to measure sensor resolution), determining a brightness value that ranges from 0 (black) to 255 (white). Every brightness value from 1 to 254 is a shade of gray—the higher a photosite’s number, the brighter its tone.
Armed with the brightness values for each photosite in the image, the camera is ready to build the image’s histogram. The horizontal axis of the histogram has 256 discrete columns (0-255), one for each possible brightness value, with the 0/black column on the far left, and the 255/white column on the far right (they don’t display as individual columns because they’re crammed so close together).
Despite millions of photosites to sample, your camera builds a new histogram for each image instantly, quickly adding each photosite’s brightness value to its corresponding column on the histogram, like stacking poker chips—the more photosites of a particular brightness value, the higher its corresponding column will spike.
Reading a histogram
The version of a picture that displays on your camera’s LCD is great for checking the composition, but the range of tones you can see in your LCD preview image varies with many factors, such as the camera’s LCD brightness setting and the amount of ambient light striking the LCD. Most important, because there’s more information in captured than the LCD preview can show even in the best conditions, you’ll never know how much recoverable data exists in the extreme shadows and highlights by relying on the LCD preview.
It’s human nature to try to expose a scene so the camera’s LCD image looks good, but an extreme dynamic range image that looks good on the LCD will likely have unusable highlights or shadows. As counterintuitive as that feels, exposing an image enough to reveal detail in the darkest shadows brightens the entire scene (not just the shadows), likely pushing the image’s highlights to unrecoverable levels. And making an image dark enough on the LCD to salvage bright highlights darkens the entire scene, all but ensuring that the darkest shadows to be too black. In fact, a properly exposed extreme dynamic range scene (a scene with both bright highlights and dark shadows, such as a sunrise or sunset) will look awful on the LCD (dark shadows and bright highlights). The histogram provides the only reliable representation of the tones you captured (or, in your live-view LCD display or mirrorless electronic viewfinder, of the tones you’re about to capture).
There’s no such thing as a “perfect” histogram shape. Rather, the histogram’s shape is determined by the distribution of light in the scene, while the left/right distribution (whether the graph is skewed to the left or right) is a function of the amount of exposure you’ve chosen to give your image. The histogram graph’s height is irrelevant—information that appears cut off at the top of the histogram just means the graph isn’t tall enough to display all the photosites possessing that tone (or range of tones).
When checking an image’s histogram for exposure, your primary concern should be to ensure that the none of the tone data is cut off on the left (lost shadows) or right (lost highlights). If your histogram appears cut-off on the left side, shadow detail is so dark that it registers as black. Conversely, if your histogram appears cut off on the right side, highlight detail is so bright that it registers as white.
Managing a histogram
In a perfect world, when you see your histogram cut off on the left (everything cut off on the left is detail-less black), you simply increase the exposure until the histogram shifts right (brighter) far enough that no shadow data is cut off. And if you see your histogram is cut off on the right, you decrease the exposure until the histogram shifts left (darker) enough that no highlight detail is cut off. Problem solved.
But many scenes contain a broader range of light, from the darkest shadows to the brightest highlights, than the camera can handle. In these scenes you can blend multiple exposures that cover the entire range of tones, apply a graduated neutral density filter (to moderate the sky). When those options aren’t available or practical, I usually save the highlights and sacrifice the shadows.
While the general goal is to ensure that none of the tone data is cut off on the left or right side of the histogram, the exposure you choose for a scene is ultimately a creative choice that isn’t bound to the way the scene looks to your eye. Though I often expose my scenes to match the amount of light my eyes see, sometimes I decide to make the scene darker or brighter than what I see.
A picture is worth a thousand words
Trusting a histogram over a picture you can actually see requires a leap of faith. I can explain the concept until I’m blue in the face, but in my workshops the point doesn’t usually hit home without a demonstration. For example, the two Horseshoe Bend sunstar images below are from the same file—on the left is the way the picture looked when I captured it, along with its histogram; on the right is the same picture with just a few minutes of very basic processing in Lightroom and Photoshop (no plugins, blending, or any other elaborate processing).
If I’d have exposed this scene bright enough for the shadows to look good on my LCD (more like my eyes saw them), the highlights would have been hopelessly overexposed (white); if I’d have darkened my highlights enough to look good on my LCD, the shadows would have darkened to an unrecoverable black. I knew my best chance for capturing this high dynamic range scene with a single click was to ignore the LCD and trust the histogram.
Despite an image that didn’t look good at all on my LCD, the histogram on my Sony a7R II showed me that I’d captured virtually all of the scene’s shadows and most of its highlights. And because I captured this image in raw mode, I was confident that I had even more shadow and highlight information than my histogram indicated, a fact instantly confirmed with a rightward tug of Lightroom’s Shadows slider. With minimal processing effort, I was able to achieve final result you see here. (This is why landscape photographers are always begging for cameras with more dynamic range, and also why I use the Sony a7RII.)
Posted on August 10, 2017
America’s National Parks have always been busy in the summer, but in recent years the summer crowds have virtually overwhelmed many of our parks. Between gridlock on the roads, more cars than parking places, and hip-to-hip tourists at the vista rails, what was once an opportunity to commune with nature has become a survival of the fittest endurance test.
My solution has been to avoid the national parks in summer, but for many summer is the only time to visit the special locations they’ve longed to see for their entire lives. And the only thing worse than visiting Yosemite or Grand Canyon in summer, is never visiting them at all.
Though I can’t make the crowds go away, let me offer an experience-based suggestion that is guaranteed to enhance your national park experience: Sunrise. Or more accurately, the morning hours from about thirty minutes before sunrise until around two hours after sunrise.
For most people the idea of rising before the sun on a vacation is laughable, but therein lies the genius. If you can overcome the urge to be most people, you can enjoy America’s most crowded national parks, at the height of the summer rush, in glorious peace. You won’t be alone, but you’ll be savoring the day’s first rays with a microscopic subset of the park’s total visitors, kindred spirits who relish nature and solitude as much as you do, who speak softly, stroll slowly, and respect personal space.
About this image
As much as I try to leave the national parks to the tourists in summer, my desire to photograph the lightning and rainbows of the Grand Canyon’s summer monsoon leaves me no choice. A couple of days ago, Don Smith and I guided our photo workshop group out to photograph sunrise at Grandview Point on the always crowded South Rim. Grandview is one of Grand Canyon’s most popular spots, but leaving our hotel about 45 minutes before sunrise got us out there about a half hour before the sun, and long before the tourists had even hit their snooze button the first time.
There were just a couple of other cars in the parking lot, the same lot that in just a few hours people will be circling in vain for five, ten, even fifteen minutes. Having Grandview virtually to ourselves, the group was able to spread out and find their own view of the canyon without competing with the teaming midday hordes that most people experience there.
Along with a few other people in the group, I set up in front a concave sandstone rock with a view across the canyon to where the sun would soon appear. Because this is my first trip with my new Sony 12-24 f/4 G lens, I’ve been making a point to familiarize myself with it, so I twisted it on and went wide. With a clear horizon and relative dearth of clouds, I dialed my f-stop to f/18 to ensure a good sunstar when the sun crested the horizon, and composed a frame.
When photographing a sunrise, the advancing light makes it impossible to set the exposure very far in advance. In these rapidly changing conditions, I love my mirrorless Sony a7RII’s pre-capture histogram in my viewfinder—I just kept my eye on the histogram, dropped the shutter speed in 1/3-stop increments as the horizon brightened, and was ready to hit the ground clicking the second the sun appeared.
Click an image for a closer look and slide show. Refresh the window to reorder the display.
Posted on May 22, 2017
When Sony asked Don Smith and me to try out their new lenses, I immediately knew where I wanted to be in Yosemite with the 12-24 f4 G lens. After great success photographing El Capitan and Half Dome as I’ve never been able to before (okay, well there was that one time last year when I borrowed a friend’s ultra-wide lens), I was ready to go home. But before leaving, I decided to walk up to the bridge beneath Lower Yosemite Fall.
With my Yosemite Moonbow and Dogwood workshop starting in just five days, my goal this morning was more to see exactly how wet it is on the bridge than it was to take any more pictures, but I decided to take my camera anyway. On the way back I played with ultra-wide (12mm) vertical compositions of this scene. Still getting used to how much I can actually fit in my frame at 12mm, I flipped the camera to horizontal and was startled to find the sun in the right corner of my viewfinder. Startled because from my location, the top of Yosemite Falls is due north (0 degrees), and the sun at that time was at 125 degrees azimuth (35 degrees south of due east).
I quickly came to terms with this revelation and repositioned myself until the sun was behind a tree, dialed to f/20, composed, metered and focused, then clicked as the sun peaked out. For the next ten minutes or so I moved as the sun moved, keeping my lens right on the edge of the shadow.
I knew the sunstar’s highlights would certainly be clipped, but I wanted to give the shadows as much light as possible without losing the highlights in the waterfall. And as important as the histogram is in these scenes with brilliant highlights and dark shadows, I knew that it wouldn’t tell me the entire story. As I increased the light by lengthening my shutter speed, in my viewfinder (I love mirrorless!) I monitored both the shadow side of the histogram and the highlight alert in the fall. I know that shooting raw, I can increase the exposure a stop beyond where the highlight alert appears, but in this case I found that I only needed to add 2/3 stop before the histogram showed me that I had all the recoverable data in the shadows I needed.
A few words about sunstars
Sunstars can be overdone, but they’re often the best way to make something interesting in difficult light. When I find myself wanting to photograph a clear sky scene facing the sun, I often use the sunstar to add visual interest to a sky that is otherwise pretty boring. Often the sunstar makes an excellent counterbalance to another strong visual object. And while a sunstar isn’t exactly what our eyes see when we look toward the sun, I think it makes a pretty good substitute for the blinding experience of looking into the sun. Take a look at the gallery of images below and ask yourself how many of these images would have been as visually appealing without a sunstar spicing up the sky.
To capture a sunstar, use a small aperture (I usually use f/16 or smaller), remove any filters (to minimize flare), and place the sun on a hard edge with most of the sun obscured: the horizon, a cloud, a tree, a flower, and so on. The more sun visible, the bigger (and more blown out) the sunstar will be. As a general rule, I try to avoid too much sun. And since each lens creates a slightly different sunstar, it helps to experiment with different lenses to determine which ones work best.
I’ll be on my annual Grand Canyon raft trip, off the grid and unable to respond, until May 30
Posted on April 19, 2015
For wildflower photography I prefer the diffuse light and soft shadows of a cloudy day, but when Mother Nature delivers clear skies and harsh sunlight, I look for backlight opportunities. Backlit flowers and leaves glow like they’ve been plugged in, and their brilliance allows faster shutter speeds that will compensate for a small aperture and quell a flower-waving breeze.
A frustrating downside of backlight is that the sun is more or less in the direction of your backlit subject, risking lens flare (scattered light that manifests as a contrast-robbing haze or distracting artifacts). If the sun isn’t in your frame, shading your lens will eliminate the lens flare. A lens hood helps, but I find lens hoods more trouble than they’re worth. Instead, when I encounter lens flare, I shade my lenses with my hand, a hat, or an umbrella (no camera bag should be without one). Or better yet, I do my best to position my lens in the shadow of a nearby tree.
But when the sun is in the frame, no amount of shading will work. In these situations I make the best of a bad thing by looking for sunstar opportunities. On last week’s visit to Columbia Hills State Park in southern Washington, I found a hillside awash with wildflowers—mostly yellow balsam root and violet lupine—in brilliant sunlight.
While waiting for the shade to arrive, I decided to take advantage of the backlight and look for a sunstar opportunity. The lupine were in better shape than most of the balsam root, and soon my eyes landed on a colorful group I could balance with Mt. Hood and the setting sun. I stopped down to f20, pulled out my Singh-Ray 2-stop graduated neutral density filter, and waited for the sun to drop to the horizon.
To salvage as much of my highlights as possible, I gave the scene as little exposure as I thought I could get away with. The foreground was pretty dark on my LCD, but the histogram looked okay (not perfect, but manageable)—in Lightroom I was able to pull up the shadows and subdue the highlights enough to work with. Photoshop’s Content Aware Fill tool helped me clean up the worst lens flare, but I still ended up with a little more than I like. (Oh well.)
Here’s my sunstar recipe (excerpted from a previous post):
Click an image for a closer look, and a slide show. Refresh the screen to reorder the display.
Posted on March 4, 2015
March 4, 2015
Just a quick update from Maui, where I’m in the midst of my annual Maui workshop (and because there’s nothing better to do when you wake up at 4 a.m. than post a blog). Before the workshop started I held my breath as I warned my group that on our first morning we’d need to leave at 3:30 to photograph sunrise from the 10,000 foot summit of Haleakala. Not only that, they should pack for temperatures in the 30s and wind. Oh yeah, and there’s a chance that the summit would be engulfed in clouds and we’d see absolutely nothing. The only defense I could offer the insane start time was that everyone needs to watch the sun rise from Haleakala at least once in their life.
Yesterday was the day. To ensure that we’d all be able to find a place for our tripods, we arrived an hour before the 6:45 sunrise. Exiting the car at the Haleakala Visitor Center vista, I looked upward, found a sky filled with stars, and immediately whispered a quick thank you to the photography gods. In addition to clear skies, relatively calm wind made the 35 degree temperature feel downright balmy.
It seems that each time I do this, Haleakala’s a little more crowded. Some in the group stayed with the masses lined along the rail at the primary view; those who didn’t mind a short but steep climb in thin air followed me a few hundred yards up a nearby trail to an elevated, less crowded view.
By the time the sun rose, I’d been playing with compositions long enough to have a pretty good idea what I wanted to do. Because of the scene’s extreme dynamic range, I decided to use my new Sony a7S, adding a Singh-Ray 3-stop reverse graduated neutral density filter, a combination that allowed me to capture this shot with a single click. I’ve never been able to capture so much foreground detail, while retaining color in the sun, in a single frame. To get the sunstar, I stopped down to f20 and clicked just as the sun peeked above the horizon.
Our sunrise success was a great start to what proved to be a wonderful (albeit long) day of photography. We descended the mountain shortly thereafter, and after a quick Starbucks recharge in Kahului, wrapped up our morning with a nice shoot in the lush Iao Valley. The afternoon included a sensor cleaning seminar, a blowhole (accented with a few whale sightings), and a cloudy but beautiful sunset on a hidden volcanic beach.
Today it’s the Road to Hana….
Click an image for a closer look, and a slide show. Refresh the screen to reorder the display.
Posted on November 11, 2014
Yosemite isn’t an inherently great sunrise location. Because most of the views in Yosemite Valley face east, not only are you looking up from the bottom of a bowl, you’re composing toward the brightest part of the sky, at the shady side of your primary subjects (Half Dome, El Capitan, Yosemite and Bridalveil Falls). This is one reason I time my workshops to include one more sunset than sunrise. But I’ve come to appreciate Yosemite Valley mornings for its opportunities to create unique images that don’t resemble the beautiful but oft duplicated afternoon and sunset pictured captured when the iconic subjects are awash with warm, late light.
Over the years I’ve accumulated a number of favorite, go-to morning spots for Yosemite. I love the first light on El Capitan, which starts at the top about 15 minutes after the “official” (flat horizon) sunrise and gradually slides down the vertical granite, is a particular treat when reflected in the shaded Merced River. Other morning favorites include pre-sunrise silhouettes from Tunnel View (especially when I can include a rising crescent moon), the deep shade of Bridalveil Creek beneath Bridalveil Fall, and winter light on Yosemite Falls.
And then there’s Cook’s Meadow. Each spring you can photograph the fresh green of the meadow’s sentinel elm, Yosemite Falls booming with peak flow, and Half Dome reflected in still, vernal pools. In winter the tree is bare, exposing the twisting outline of its robust branches. The highlight each autumn is the few days when the tree is bathed in gold. On the chilliest fall mornings, sparkling hoarfrost often decorates the mounded meadow grass, and if you’re really lucky, when the air is most still, you’ll find the meadow hugged by an ephemeral mist that rises, falls, disappears, and reappears before your eyes.
On last week’s workshop’s opening morning, after a nice sunrise silhouette shoot at Tunnel View, I rushed my workshop group to Cook’s Meadow in time for the first light there. We hit the autumn big three: a hoarfrost blanket, the elm’s autumn gold still going strong, and even a few wisps of mist. The image here I captured toward the end of our shoot, just as the sun kissed the valley floor. My wide, horizontal composition emphasized the foreground, which was far more interesting than the bland (and contrail scarred) sky. I dialed in a small aperture to enhance the sunstar effect, and used a Singh-Ray 2-stop hard-transition neutral density filter to moderate the bright sky.
Within minutes the light was flat and the mist was gone, but the group was happy. Not a bad start to what turned out to be a great week of photography.
Click an image for a closer view, and to enjoy the slide show
Posted on August 25, 2014
A few days ago I was thumbing through an old issue of “Outdoor Photographer” magazine and came across an article on Lightroom processing. It started with the words:
“Being able to affect one part of the image compared to another, such as balancing the brightness of a photograph so the scene looks more like the way we saw it rather than being restricted by the artificial limitations of the camera and film is the major reason why photographers like Ansel Adams and LIFE photographer W. Eugene Smith spent so much time in the darkroom.” (The underscores are mine.) Wow, this statement is so far off base that I hardly know where to begin. But because I imagine the perpetuation of this myth must send Ansel Adams rolling over in his grave, I’ll start by quoting the Master himself:
Do those sound like the thoughts of someone lamenting the camera’s “artificial limitations” and its inability to duplicate the world the “way we saw it”? Take a look at just a few of Ansel Adams’ images and ask yourself how many duplicate the world as we see it: nearly black skies, exaggerated shadows and/or highlights, and skewed perspectives. And no color! (Not to mention the fact that an image is a two-dimensional approximation of a three-dimensional world.) Ansel Adams wasn’t trying to replicate scenes more like he saw them, he was trying to use his camera’s unique (not “artificial”) vision to show us aspects of the world we miss or fail to appreciate.
You’ve heard me say this before
The rest of the OP article contained solid, practical information for anyone wanting to come closer to replicating Ansel Adams’ traditional darkroom techniques in the contemporary digital darkroom. But it’s the perpetuation of the idea that photographers are obligated to photograph the world like they saw it that continues to baffle me.
The camera’s vision isn’t artificial, it’s different. To try to force images to be more human-like is to deny the camera’s ability to expand viewers’ perception of the world. Limited dynamic range allows us to emphasize shapes that get lost in the clutter of human vision; a narrow range of focus can guide the eye and draw attention to particular elements of interest and away from distractions; the ability to accumulate light in a single frame exposes color and detail hidden by darkness, and conveys motion in a static medium.
No, this isn’t the way it looked when I was there
While this sunset scene from Lipan Point at the Grand Canyon is more literal than many of my images, it’s not what my eyes saw. To emphasize the solitude of the lone tree, I allowed the shaded canyon to go darker than my eyes saw it. This was possible because a camera couldn’t capture enough light to reveal the shadows without completely obliterating the bright sky (rather than blending multiple images, I stacked Singh-Ray three- and two-stop hard transition graduated neutral density filters to subdue the bright sky).
To convey a mood more consistent with the feeling of precarious isolation of this weather-worn tree, I exposed the scene a little darker than my experience of the moment. The sunstar, which isn’t seen by the human eye but was indeed my camera/lens’ “reality” (given the settings I chose), was another creative choice. Not only does it introduce a ray of hope to an otherwise brooding scene, without the sunstar the top half of the scene would have been too bland for me to include as much of the shadowed canyon as I wanted to.
I’m not trying to pass this image off as a masterpiece (nor am I comparing myself to Ansel Adams), I’m simply trying to illustrate the importance of deviating from human reality when the goal is an evocative, artistic image. Much as music sets the mood in a movie without being an actual part of the scene, a photographer’s handling of light, focus, and other qualities that deviate from human vision play a significant role in the image’s impact.
(Stuff my camera saw that I didn’t)