I've been pleased with my recently purchased 35L except for the CA in the (full frame) corners. Much worse than any other lens I own but not really noticeable unless pixel peeping so I'm happy enough to keep the lens as long as this is normal.
Yep, the 35L gives you CA hell... especially in the corners when using FF. It doesn't go away with stopping down either. On top of that it is very difficult to correct it (I haven't managed to get rid off it fully in PP). My 24-70L has way less CA's and is 100% correctable in PP. So, based on my own experience I would say this is normal.
Most of the time I use this lens with wide apertures. CA in the corners blends away in the blurred background. So in these cases it is not much of an issue for me. It is most noticeble when shooting contrasty landscapes...
PetKal wrote:
Yeah, but wide open, and up to f/2.2 or so on mine....where it just about disappears. Not at f/4.
That f/4 CA surprises me. I'll check my 35L today for that.
The (purple) fringes in the center of the frame are gone when stopped down to f/2 or f/2.8. But these fringes are not the same kind of CA's that are in the (extreme) corners of the frame (and to which the OP is referring to).
If you look at the color fringing on the 24~105 image, it has the usual cyan on one edge, red on the other. It is known as Transverse Chromatic Aberration*. None in the center, and gets progressively worse the more you move towards the edges of the image. Easily corrected by all the common software tools (that simply resize one of the color channels to re-register all three colors).
The 35L image does not show that type, but instead shows a hard purplish edge on just one side. Often referred to as the dreaded purple fringing. Technically known as Longitudinal Chromatic Aberration*. It is most visible in areas of strong contrast/backlighting, etc. The typical software tools will have no effect here, as it can happen anywhere in the image. Here is one action that works well to remove it. http://www.shaystephens.com/#ps
jjlphoto wrote:
If you look at the color fringing on the 24~105 image, it has the usual cyan on one edge, red on the other. It is known as Transverse Chromatic Aberration*. None in the center, and gets progressively worse the more you move towards the edges of the image. Easily corrected by all the common software tools (that simply resize one of the color channels to re-register all three colors).
The 35L image does not show that type, but instead shows a hard purplish edge on just one side. Often referred to as the dreaded purple fringing. Tecnically known as Longitudinal Chromatic Aberration*. It is most visible in areas of strong contrast/backlighting, etc. Harder to remove, as it can happen anywhere in the image. Here is one action that works well to remove it. http://www.shaystephens.com/#ps
A further aspect of Longitudinal CA that is not discussed by van Walree is its presence in the OOF regions of the image. Photozone.de often tests for this and shows images like this:
jjlphoto wrote:
If you look at the color fringing on the 24~105 image, it has the usual cyan on one edge, red on the other. It is known as Transverse Chromatic Aberration*. None in the center, and gets progressively worse the more you move towards the edges of the image. Easily corrected by all the common software tools (that simply resize one of the color channels to re-register all three colors).
The 35L image does not show that type, but instead shows a hard purplish edge on just one side. Often referred to as the dreaded purple fringing. Tecnically known as Longitudinal Chromatic Aberration*. It is most visible in areas of strong contrast/backlighting, etc. Harder to remove, as it can happen anywhere in the image. Here is one action that works well to remove it. http://www.shaystephens.com/#ps
A further aspect of Longitudinal CA that is not discussed by van Walree is its presence in the OOF regions of the image. Photozone.de often tests for this and shows images like this:
All the fast L primes act like this (with perhaps the 200/2 being the exception), it is one area they don't shine, but they tend to be better than their non L counterparts....The 35L is one of Canons finest, just use it and enjoy the goodness
Thanks everyone for your terrific input, especially the great technical posts.
From previously reading I had already expected some CA out of the 35L but was surprised by the amount, especially compared to the my other lenses. I didn't know if it was normal.
This shot was deliberately aimed at showing the aberration and I doubt I'll be seeing it in regular shooting. Even when it does appear I probably won't even bother trying to deal with it because it's hardly noticeable. Now that I've received reassurance (from other owners) that my 35L is on par I'll stop testing and just use it. It really is very pleasing in every other way.
You will also see this with German glass (Zeiss, Leica). It is essentially a by-product of fine lens design. It is possible to design a lens without this phenomena, but then the trade off for very highly corrected axial/longitudinal aberration can be often be overly harsh or extremely busy bokeh.
jjlphoto wrote:
You will also see this with German glass (Zeiss, Leica). It is essentially a by-product of fine lens design. It is possible to design a lens without this phenomena, but then the trade off for very highly corrected axial/longitudinal aberration is overly harsh or extremely busy bokeh.
No worries ... the trade-off is fine by me. Interesting about the German glass as I'm keen to see how the new Zeiss 21mm pans out. I fancy a quality ultra-wide too.
