Those ISO 12233 Crops from Nikon lenses

[cosmo.6792]

I noticed a complete lack of vignetting or chromatic aberration in every single one of the ISO 12233 crops from Nikon lenses. I suspect that the operator of this website wasn

[Jon Ruyle]

I see plenty of CA ("http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?FLI=0&API=0&Sample=0&FLIComp=0&APIComp= 0&LensComp=0&CameraComp=0&SampleComp=0&Lens=646) in the ISO 12233 crops of the Nikons.


Bryan's ISO crops of the canon lenses are extremely well done. I don't know for sure, but I'd be shocked if he was posting crops with optional correction turned on. (Of course, there is some amount of cooking of the raw that cannot be turned off and we can't do anything about that).

[cosmo.6792]

Jon,


The sample of CA you posted is the Purple Fringing ("sensor blooming") type. That's the type that occurs with very large aperture lenses shot wide open. It's visually characterized by color aberrations in all directions by a single color (i.e. Purple). The effects are strong across the entire frame, especially in the center. Cameras and RAW converters do not possess the ability to automatically correct that type of CA. So it's no surprise that the Nikon crops exhibit this flaw.


The type of CA that I'm talking about is Lateral CA. These aberrations occur in two symmetric directions, usually in TWO colors (i.e. magenta & cyan). The effects of Lateral CA is rarely seen in center of the frame, but it gets progressively worse towards the far edges of the frame. This is the type of CA that can be automatically corrected by Nikon cameras and/or RAW processors. The total lack of CA (and vignetting) in all ISO 12233 crops from Nikon leads me to suspect that Bryan forgot to deactivate the automatic correction during RAW conversion.


Here is an example ("http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?FLI=0&API=0&Sample=0&FLIComp=0&APIComp= 0&LensComp=646&CameraComp=614&SampleComp=0&Lens=28 6) of the difference between the two. Notice the cyan and magenta aberrations in the corner crops. That's the lateral CA i'm talking about. Mouse over the sample, and you'll see the other type of CA, which is sensor blooming (or purple fringing). One is correctable, the other is not.

[Jon Ruyle]

I noticed a complete lack of vignetting or
chromatic aberration in every single one of the ISO 12233 crops from
Nikon lenses.


So by "complete lack of CA" you meant "complete lack of lateral CA".


I just reread Bryan's description of how the ISO12233 crops are produced, and saw nothing about CA. As I said, I'd be surprised if he overlooked something this simple, but I really don't know. If Bryan or someone else who knows the answer does not respond in this thread, you might try emailing him directly with your question.

[neuroanatomist]

The sample of CA you posted is the Purple Fringing ("sensor blooming") type. That's the type that occurs with very large aperture lenses shot wide open. It's visually characterized by color aberrations in all directions by a single color (i.e. Purple).


Purple fringing is more properly known as longitudinal (or axial) CA. Sensor blooming is not accurate - there is no color restriction to a sensor bloom, nor would a sensor bloom change from purple to green depending on the direction of defocus. You can see longitudinal CA on film, too - it's not a phenomenon of the sensor.


Longitudinal CA results from failure of a lens to focus all the wavelengths of light at the focal plane. When the green light is properly focused (which is usually the case), longer wavelengths (red) are focused behind the sensor plane, and shorter wavelengths (blue) are focused in front of the focal plane. The OOF red and blue light 'sum' to magenta, and that's what gives rise to the purple fringing.


But I have to say 'wow' - the fast Nikon lenses seem to have huge amounts of longitudinal CA, e.g. compare the Canon 135mm f/2 with the Nikon 135mm f/2. However, looking at other sites which conduct objective testing, the differences in amount of LoCA seem much less pronounced.

[wickerprints]

The OP is correct. The ISO 12233 crops from Nikon have certainly been corrected for lateral CA, because every single chart from every Nikon lens shows no such aberrations, which is impossible. For proof all you have to do is pick an ultra wide-angle lens and look at the f/8 result. Since lateral CA is not reduced by stopping down, but axial CA ("purple fringing") is, stopping down allows you to tease the two effects apart. Nikon glass is NOT so good that even their UWA lenses are free of CA; otherwise they would

[neuroanatomist]

Oh, and about that axial CA: Nikon supertelephoto lenses are very purple wide open, which is not too surprising, because they don't use pure fluorite elements. Thus the efficiency of their CA correction is limited.


Thanks, that makes sense. But even at shorter focal lengths (85mm, 135mm) the Nikon lenses show substantial LoCA, whereas the Canon equivalents do not. Looks like even in the absence of fluorite elements, the UD elements do a pretty good job at correcting the LoCA.

[Jon Ruyle]

Nikon supertelephoto lenses are very purple wide open, which is not too surprising, because they don't use pure fluorite elements. Thus the efficiency of their CA correction is limited.


I don't know too much about exotic glass, but he best ED telescopes are every bit as good as the fluorite ones. Telescopes are slower, but the demands on color correction are probably more stringent.