Terrywoodenpic wrote in post #5386471
Purple fringing certainly is caused by the sensor micro lenses.... but the more Telecentric the camera lens is, the less the effect there is likely to be.
The problem is caused when the light from your lens strikes the micro lens at acute angles. Leica and some other cameras offset the microlenses so that light passes more directly to the sensor, and reduces fringing. It like CA is usually more of a problem with poor quality or wide angle lenses. ( i.e. not telecentric).
Whilst CA is found equally on Film cameras, Purple fringing is not.
I was buying this explanation for a while, but the more that I think about it, the less plausible it seems. The purple fringing that I have observed while using cheap lenses occurs at all points on the sensor, not just the outer edges. It also appears to be much more prevalent when the lens exhibits a large amount of flare. The problem is the worst on my cheap 75-300mm telephoto lens as opposed to the EF-S 18-55mm f/3.5-5.6 kit lens. Further, the purple fringing problem also occurs on cameras whose sensors do not have micro lenses. I have read other descriptions of the problem that attribute the cause to saturation of sensor elements and, indeed, that is where the problem seems to be the most noticeable.
While it is easy to envision that the very rudimentarymicro lenses can suffer from chromatic aberration, purple fringing seems to uniformly surround bright areas that are near saturation rather than behaving as CA where complementary diffraction components exist on opposite sides of high contrast areas. It is also interesting to note that the width of the purple fringe seems to be dependent upon the brightness of the light which is counter to the idea of being diffraction related.
I won't completely dismiss the idea of color fringing due to the micro lenses (after all, it is glass), but for the time being and without a better and more rigorous explanation, the idea that it is always purple doesn't quite seem to float.
