As the next generation of Nikon and Canon cameras hit the street, one of the questions that always occurs to me is how these new bodies interpret color, in relation to the tried and true models they are replacing. I will be testing more than one such pair, but lets begin with the newly released Canon 5D Mark lll, and its predecessor the Mark ll.
There are superficial changes between the cameras which remind you that they are not the same body, such as the on/off switch being moved to a new location, the “on but not adjustable” notch on the on/off switch being removed, grippy material being added where the heel of the hand rests on the body, and the larger LCD display. The change I will most appreciate, from a mechanical point of view, is that the mode dial can no longer be turned by accident, meaning I will not need to put a piece of black electrical tape on the top of my new Mark lll, as was needed on the Mark ll, to assure that the mode was not accidentally changed. Others will list the pros and cons of the various improvements, which appear to offer more for the video shooter than the still photographer. I will attempt to focus solely on their color reproduction.
In order to do this a copy of the Datacolor SpyderCheckr color target was set up and lit in a reasonable manner, with a single light from a 45 degree angle. It was shot at somewhat less than full frame to eliminate the corners of the image, where the lens is most challenged, from the target. In front of this a tripod was located and configured. Each camera was set, with the same lens, to full manual settings, using the same aperture, shutter speed, and ISO, at the same focal length on the lens. Both were set to shoot full size RAW images, and both produce similarly sized resulting files. So most factors are matched between the two cameras, so that differences in the resulting files should be representative of actual differences in the flavor of color each produces.
Multiple exposures were shot with each body, and the sharpest chosen from each. Sharpness has little effect on averaged color values from large patches, but its the habit of a lifetime to select the best files, so it was done here as well.
Since neither Lightroom 4, nor the Photoshop 6 beta, even with the latest ACR plugin from Adobe Labs will open the RAW files from the new Mark lll, it was necessary to use the Adobe DNG Converter, v 6.7 beta from Adobe Labs to process the files. Files from both cameras were processed this way for consistency. Both resulting DNG files were then opened in Lightrooom 4, and processed as any SpyderCheckr target shot is processed, by balancing on a light gray patch, adjusting the exposure to get appropriate, and similar white values from both shots, and adjusting blacks to get appropriate and similar black values as well.
The two resulting adjusted Lightroom files were then exported to the SpyderCheckr software, where a calibration set was produced using the Colorimetric option. Both resulting presets were then selected in Lightroom, and screenshots of the HSL adjustments they contained were screenshot. In the image below, the adjustment set on the left, as the title notes, is the 5D Mark ll, while the set on the right represents the Mark lll. While the HSL adjustment units in Lightroom are not as absolute as Delta-e comparisons, they are quite straightforward, and offer at least a relative sense of the color corrections required by each body, using the same lens.
Numerically, its possible to come up with some averages of the adjustments that SpyderCheckr made to correct the color of both bodies. The total of the Hue adjustments made to the eight color channels was 35 for the Mark ll, and 38 for the Mark lll. Saturation adjustments were 55 for the Mark ll, and 60 for the Mark lll. And Luminance adjustments were a total of 4 for both bodies. So the numerical route would tend to indicate quite similar color accuracy between the two cameras. Now what about the location and degree of those corrections?
The Mark ll scored the largest of all adjustments, a -19 in Green Hue. The Mark lll also scored its largest adjustment on the same setting, but only a -16, instead of -19. Similarly the second and third largest Hue adjustments were shared by both bodies as well, with the Mark ll requiring a +4 on Red and -4 on Aqua Hues, while the Mark lll scored +6 and -4 on the same two Hue channels. The outlier here is the Mark lll requiring 7 points more reduction in Magenta than the Mark ll.
The two bodies show similar flavor on Saturation results as well, with Saturation decreases on nearly all colors, reinforcing the theory that Canon’s produce more saturated color than Nikons, but you’ll need to wait until a further article in this series to see a cross-comparison that addresses this. The channels requiring the largest desaturation are similar in both bodies as well, with Red, Blue, and Aqua requiring larger adjustments with both, while the Mark lll reduced the size of the Green adjustment, but increased the size of the Magenta adjustment needed.
The Luminance variations in both cameras are very small, with a maximum of -3 in the Mark ll, and of +2 in the Mark lll. Overall, the luminance balance is excellent, barely requiring any tuning.
In general, what color effects should one expect to see from these two cameras, when uncalibrated? By the numbers the over-saturation in Reds and Greens would be the most obvious in the Mark ll, and the over-saturation in Red and Magenta in the Mark lll. Perhaps the easiest way to tell them apart would be the Green Saturation levels, with ten points more saturation in the Mark ll than the Mark lll. Keep in mind that building a calibration for each body, and applying it to all the images shot with that camera should go a long way towards removing any visible color errors in either case, and should allow excellent matching between their files, in case you now use your Mark ll as your second body, and need to mix images shot with both.
The image below is an animated GIF showing the normalized, but not color corrected, results from each camera. The areas of difference noted above do stand out, but the blues appear, visually, to have more variation than any other color.
It may be necessary to click on this illustration to bring it up in its own window in order for the GIF animation to occur.
A Note on Camera Color
What cabbage leave does camera color come out from under? It has two parents: the camera manufacturer, who is responsible for the hardware, and for the firmware that produces the proprietary file, and the RAW converter developer, who uses their own form of camera profiles to convert that proprietary camera format to a RAW file on screen. At one time, both these were the same parent; and we felt obliged to use the manufacturer’s RAW converter when testing the camera. But today the majority of RAW files, at least those from DSLR and subDSLR cameras, are processed in third party software. And by far the most common of those third party products is the Adobe engine used in Lightroom, ACR, and therefore Photoshop. So when I talk about camera color, I’m speaking of the camera color we are going to get in the real world, where we process our images in Lightroom or ACR. Using another third party RAW converter, or a camera manufacturer’s conversion utility, may produce different results. But given the success I have been having with the Adobe engine and a SpyderCheckr HSL correction set, I am quite comfortable with the color from this workflow, and don’t see any significant reason to be looking at less common or more complex workflows in order to achieve the results I need; especially with Lightroom 4’s much improved capabilities.
Please read my additional article on Canon 5D Mark lll color.