Measure Axial Chromatic Aberration: MTF Mapper Part Deux
This article will show you how to use the mtf_mapper_gui.exe program to measure axial (longitudinal) chromatic aberrations. Longitudinal chromatic aberration (LoCA) is the optical problem of focusing different colors of light at different distances along the optical axis.
My other article about MTF Mapper is located here . The other article is about MTF50 resolution measurement and focus calibration. Be aware that you need to use separate charts to make separate measurements (lens MTF50 resolution, focus calibration, or LoCA analysis). Also be aware that measurements depend upon the color of light used while photographing the targets (I used outdoor lighting with a clear sky in the sun).
LoCA differs from lateral (transverse) chromatic aberration, which instead spreads out different colors perpendicular to the lens optical axis. Lateral chromatic aberration typically shows up as purple corners in the photograph; you won’t see it in the center of the image. Lateral chromatic aberration is simple to fix; most modern cameras can even automatically fix it in-camera. Longitudinal chromatic aberration is more difficult to correct, although programs such as Nikon Capture NX2 can largely mask its effects.
Axial (longitudinal) aberration diagram, courtesy of Wikipedia.org
LoCA can rob a lens of resolution, since some colors will be in-focus and some colors will be out of focus. You can recover the resolution by stopping the lens down until all visible wavelengths are in focus, but artistically this is often a poor option.
A common visual effect of LoCA is seeing magenta instead of white in specular reflections on the eye.
If you’re interested in evaluating how much LoCA a lens has, then the MTF Mapper program can help you measure it, and the program is free at the time of this writing.
The MTF Mapper program author is Frans van den Bergh. His software and printable test charts are available here. Frans comes across as a ‘brainiac’, and his documents can take your breath away. They’re worth a read, though, even if you can’t grok 100% of their contents. More of his writings about image analysis topics can be found here. If you like his stuff as much as I do, please let him know! I’ll bet Frans has spent a gazillion hours working on this software, and he deserves all the praise we can give him.
I’m using version 0.5.7 of mtf_mapper_gui.exe for these tests. You need to print the proper chart and use the correct program preferences to get the desired measurements. The chart I used is called “mfperspective_a3.pdf” (printed on A3 paper), which is also available at the same site that you download the program. You’re supposed to rotate the chart 45 degrees about the vertical, with the taller vertical targets farther from the camera than the shorter ones.
Mfperspective_a3.pdf chart photo. Chart rotated 45 degrees.
The MTF Mapper program takes advantage of the fact that camera sensors are separated into red, green, and blue-sensitive pixels. The program can independently analyze single-color pixels or all of them together. By the way, half of all of the camera sensor pixels are sensitive to green, while 25% are sensitive to red and the last 25% are sensitive to blue. It turns out that human eyes are most sensitive to green, so having 50% of the pixels being sensitive to green makes pictures look ‘correct’. This kind of sensor design is called “Bayer”, named after the inventor Bryce Bayer who used to work for Kodak. Different combinations of the R,G,B pixels can recreate all of the colors we see.
MTF Mapper uses a program called “dcraw” (included in download) that knows most camera “raw” formats, and is regularly updated for new camera models.
Axial R,G,B Focus Measurement
In the following test, I’m using the Sigma 150-600mm zoom. Big lenses typically have the most trouble with axial chromatic aberration, and they also tend to exaggerate any focus errors.
Take a photo of the chart, aligning the camera focus sensor on the middle of the chart. Don’t worry if the focus isn’t perfect; what counts is the difference between the red, green, and blue color channels in the same photo. A perfect lens would focus all three colors at exactly the same distance. The largest aberration errors will be seen with the lens aperture wide open. You’ll likely get different measurement results as you change focal lengths, too.
Set your camera for RAW mode, so that there are no in-camera compensations for sharpening or chromatic aberrations. Remember to rotate the chart so that it’s at about 45 degrees from the sensor, with the tall side of the chart target images farther away than the short images.
MTF Mapper Settings Preferences to measure axial aberrations
After getting a photo of the test chart, the program needs to be configured for the color to be analyzed and for the camera pixel dimensions. The Nikon D610 has pixels of 5.95 microns, but the D7100 has 3.92 micron pixels, for instance. The program needs to be run against the same picture three times, each time selecting a different Bayer color channel.
Blue Bayer Channel focus results show focus is 14mm in front of the chart center.
Chart close-up (green). The orange arrows are at the chart center.
Green Bayer Channel shows focus is 5mm beyond the chart center.
Red Bayer Channel shows focus is 4mm beyond the chart center.
The results above indicate that the red and green colors are focused at almost exactly the same distance, but the blue channel shows focus is nearly 19mm closer than the green and red channels (at a focus distance of about 6 meters).
The test results also show the MTF50 resolution, measured in cycles per pixel. The green channel is sharpest at .211 c/p (peak sharpness). The red and blue channels both measure about 0.197 c/p. This MTF50 measurement is only valid for the peak focus location; you should be analyzing resolution using the resolution chart (see the other MTF Mapper Cliffs Notes article).
Evaluating the MTF50 Math
MTF50 lp/mm = (c/p)* pixels tall / height_mm
Line pairs/pixel height (lp/ph) = MTF50 lp/mm * sensor height mm.
D7100: 3.92 micron pixels
4000 X 6000 pixel sensor (24 MP)
15.6mm X 23.5 mm sensor
Green MTF50 = 0.211 c/p = .211*4000/15.6 = 54 lp/mm, or 844 lp/ph
Blue MTF50 = 0.197 c/p = .197*4000/15.6 = 50.5 lp/mm, or 788 lp/ph
Red MTF50 = 0.197 c/p = .197*4000/15.6 = 50.5 lp/mm, or 788 lp/ph
Conclusion
The MTF Mapper program is a great way to evaluate axial chromatic aberrations. This optical aberration is typically a bit mysterious to get a handle on, but now there’s an easy way to characterize it. When you can analyze by the numbers, then you can actually compare it against other lenses in a meaningful way. And you can’t beat the price.
