ColorChecker Passport – Matching Color Response Under Different Light Sources
by eduardoangel – November 9th, 2009
by Joe Brady
I’ve discussed the how’s and why’s of creating and using camera profiles in previous posts. One of the great benefits of having a DNG camera profile is the ability to match the color response or your camera under light sources with different color temperatures and spectral output. You’ve probably seen how your inkjet prints look different under different lights. Most lights are strong in certain colors of the spectrum and weak in others. For example, standard fluorescent tubes often have a strong green portion in their light output. Since the opposite of green is magenta, any parts of your print that have a strong magenta component (mostly reds and purples) will start to be cancelled by the green light from the fluorescent light. Conversely, standard tungsten lights are very yellow. Since blue is the opposite of yellow, blue parts of the image will lose their saturation.
Even though we can white balance under each light source, that is not enough to bring back the full saturation of each color. Let’s take a look at a few examples taken under very different lighting conditions. In each case the camera was set on daylight white balance assuming white balance would be performed in software later. As expected, the color temperature of each scene is wildly different. Next to each image is the same shot with a white balance applied using the ColorChecker Passport neutral patch. By the way, my sincere thanks to Sarah at the MAC Group for patiently modeling for us!
Each image now has a correct white balance applied and the results are certainly more pleasing. We’ll soon see however that while white balance gets you a lot closer, applying a camera profile to each image does a much better job of matching the color.
You may have noticed that purples and magentas can be very difficult colors to match as they seem to very affected by different temperatures of light. In order to adjust for that, I created a DNG profile for each lighting condition I encountered this day – one each for tungsten, daylight, shade and fluorescent light. After applying a white balance to each image, I applied the appropriate DNG profile to each image. The results can be seen below. From left to right, we have photos captured under tungsten, daylight, shade and fluorescent light sources. Notice that in the set of images where only white balance was applied, though fairly close, the color response of both the purple vest and the skin tones vary from image to image.
Applying the profile to each white balanced image however gives us much more consistent and uniform color. You may only see a subtle difference to see on your monitor, but if you apply this workflow to your images, the changes will be clear.
This may seem like a minor point, but take a moment to realize what we have just accomplished – we have matched the color response in our camera under four very different lighting conditions with just two clicks! One for white balance, and one to apply the DNG profile. That saves an incredible amount of time on image editing, eliminating the task of trying to correct each color.
In my next post, I’ll discuss matching the color response between different camera models – even when from different manufacturers!