In order to be more accurate with my measurement of chromatic aberration in my lens tests, I have changed the way that I do things. I now measure the CA at a more optimum site (more work!) and I have changed the number that I report.
I used to report CA in pixels (actually an area in square pixels), but have since learned that since the CA worsens continuously as you move from the center to the edge, the CA will vary depending upon where you measure it. The area in pixels can be corrected for the distance that the measurement is done from the center of the image. The number is reported in %.
The numbers are as follows:
- less than 0.04% = insignificant
- 0.04 – 0.08% = mild, difficult to see
- 0.08 – 0.15% = moderate, somewhat visible
- more than 0.15% = severe, easily seen
The measurement above predominantly applies to lateral CA. Longitudinal CA is more difficult to measure. Most lenses will show a little long CA at large apertures. It can be seen most prominently in out of focus portions of the image. You will commonly see a red fringe on one side of the focus plane and a cyan fringe on the other side. I use these fringes in lens testing to tell me when I am getting close to the plane of sharp focus (the fringes tend to disappear when the object is in focus).
To show long CA, I have added an image of my test pattern when it is tilted across the image, showing the focus plane in the center and getting progressively more fuzzy as you move outward. This should be show the fringes if they are present (we’ll see how it works out in real life testing).
I have also put up a new lens test for the Nikon Printing-Nikkor 105mm f/2.8 lens. This was an unusual and interesting lens to test.
