Editor’s Note: A special thanks to our partner Datacolor for helping us to bring more information about color calibration to you.
A photographer walks into a bar…
A couple of art directors are chatting over their beverages. The photographer, sitting nearby overhears their conversation. One of them says “Did you see those pictures we got yesterday? They were all dark.”
The other art director replies “Not a problem. I just made my monitor brighter…”
The photographer shakes his head in sadness. He leans over then asks, “You guys calibrate your monitors regularly don’t you?”
“Um. Er. Well… No, not really.” the art directors answered.
“No. Not really” really means “No, not ever.” Sadly, that’s the answer for most computer users today. Without color calibration, the resulting brightness and faithful color reproduction are at best a game of chance with so many variables the outcomes are mostly dark and rarely right.
What is color calibration?
Simply, color calibration is that each device in the chain from capture to editing, to proofing, to reproduction, understands the colors it receives before it passes them to the next device. Each link in the chain is an opportunity to have great color or to muck it up. As every calibrated device passes great color to the next, uncalibrated ones pass on muck that can get worse as it goes through each unit. The promise of color calibration is the accurate color from capture to post production to output. First, here’s some background on how the digital world sees brightness and color.
Neutralizing color casts
Color calibration starts at the camera. Cameras have white balance controls that get the photo in the ball park of sort of accurate color. Daylight, Cloudy, Open Shade, Tungsten or Incandescent, Fluorescent and Flash are the common choices. There are many factors that can affect the color recorded in the capture. Snow on a sunny day is always blue because it reflects the sky above. The front of a bridal gown is often an unpleasant shade of green when she is photographed on a beautiful lawn. These are examples of color cast. The color “white” changes with the light. No matter what color of light we might find ourselves, our brains quickly adjust so that white looks white no matter what. The camera records blue snow in the photo below. What I saw at the time (I swear) was white snow.
What is real versus what we think we see
The brain automatically balances color to neutral. Neutral means no color cast. In the photo above my brain told me “That snow is white.” I dutifully believed it. The camera, on the other hand, doesn’t have a brain to do this work. To dispel a notion, Auto White Balance is not your brain. Please don’t use it. Look at the photo below of three white mice. The one on the left is lit with a tungsten bulb. The one in the middle is lit with daylight and the one on the right was open shade. My brain saw three white mice. My camera does not share the bias and reported the colors accurately. A colorimetrically neutral reference is needed to make the corrections to get rid of the color cast. It’s really easy. One click in Photoshop’s Camera Raw or Lightroom’s Develop module does the trick. A pair of color reference charts is coming up. First, we have to understand exposure and brightness
Exposure & brightness
Brightness is the dark and the light of a photograph. There can be so much of it that light areas in a photo have no detail. They are completely white. There can be so little of it that nothing can be seen. These are the extremes. The balance is detail in both the lightest areas (highlights) and the dark areas (shadows.) The two main photo editing software packages Lightroom and Photoshop work the same way. Adobe has chosen to give them separate scales.
Digital photographs express brightness by the numbers. Photoshop’s numbers are the most common so I’ll use them in this post. The conversions for the numbers from Photoshop to Lightroom are shown above. Zero (0) is black. 255 is pure white. The other 254 numbers are shades of gray in this example. There are two key numbers that target brightness thresholds when making photographs. Shadows must have numbers higher than 25 to maintain detail. Highlights must fall between 242 and 248 to hold detail. Brightness numbers below 25 reproduce as black. Brightness numbers above 242-248 reproduce as white.
The combination of ISO, shutter speed and aperture must return a brightness number between 242-248 on the highlight patch of a calibrated reference chart. In my case, it’s the X-Rite ColorChecker Passport Photo. I’ve been using ColorCheckers since 1981 to calibrate the color of film and moving right on to digital capture. Another good one is the Datacolor SpyderCheckr. These reference charts are manufactured to return a specific brightness and are calibrated to do just that. Most photographers “eyeball” exposures by viewing results on the monitor on the camera. The problem with this approach is that monitor isn’t calibrated and neither is the photographer’s eye! The left photo below is what I “saw” the correct exposure to be. It looked great until I read the numbers only to discover that the photo was over a stop underexposed. I read the numbers on a ColorChecker Gray Scale chart (inset.) Then I adjusted the camera’s exposure and checked again. This is an easy way to get a perfect exposure quickly especially when shooting tethered.
Brightness and its relation to a correct exposure bring up the subject of how digital creates color. Each shade of gray has a number: 1 through 254 with black as 0 and white being 255 for a total of 256 distinctly separate tones. Perfect for a mono color world. Our world is colorful. Digital sensors record those colors with three primary colors: Red, Green, and Blue abbreviated as RGB.
How digital color works
When the RGB numbers match there is no color cast. For example, the first patch below is white. RG&B equal 255. That, of course, is white. When R&G are 255 with B at 0 the result is yellow. The chart shows the primaries and complementary colors along with a random mix. The possible combinations of colors are well over sixteen million. Vary the value for R or G or B by just one point and you have a different color.
Removing color cast
Before checking the exposure, a click on the third patch from the white one removes color casts. This photo shows the before color neutralization on the right and after the one click correction on the third patch starting with white. The right side shows R:223, G: 218, B: 206. This suggests a reddish-yellow color cast. On the neutralized left side the numbers are all the same. R:217, G:217, B:217. When the RGB numbers are the same there is no color cast. The color of the image is considered to be neutral.
Unless the photographer is willing to understand how the numbers work and what they mean, monitor calibration is an absolute necessity. This is even truer if a color reference chart is not used. As we’ve seen, our brains compensate when something is amiss color or brightness wise. Datacolor’s newly released Spyder 5 is a quick easy and affordable way to calibrate a monitor. Remeber the photographer who walked into a bar? He understands that even if the monitor’s colors are correct; if its brightness is too high or too low, anyone judging the quality of an image on it would be guessing. The only way to visually trust anything the monitor displays is if it is properly calibrated. If a photograph looks good on a calibrated monitor, it will look good on the web if its color space is sRGB.
Every printer has its own unique way of translating color it receives into the inks it puts on the paper. This is the last step of color calibration. A profile is created by printing out swatches on the printer being profiled. Those swatches are read with a device that measures the color on the paper. The software compares the paper reading to the actual color sent to the printer for each patch. The result of that comparison is called an ICC profile. When applied to the photo during printing it assures the colors in the file are mapped to the correct combination of inks.
Regular color calibration of monitors is important. So is maintaining color calibration throughout the entire process. With color calibration in place, everything turns out the way it’s supposed to look, from capture in the camera to output on the web and in print. Best of all, when your monitor is calibrated, you don’t have to deal with all of those numbers.