Normal Fringing on the 17-40?

[canoli]

Can you guys take a look at these, tell me if this amount of red fringing is normal?


Sorry if these aren't resized properly - not sure if TDP is cropping them or what, but the Preview of this post doesn't look too good. But hopefully you can what I'm talking about.


The shot was taken last night over the Hudson at sunset, directly into the afterglow.


40D + 17-40mm f/4L, tripod, f/11 @ 1.3s ISO 100, 17mm - no filter, Adobe RGB, Auto WB.


The crops are taken from just left of center. One has a CA adjustment (in ACR) - Red/Cyan to -63 (toward Red), the other is as shot.


As you can see, the red fringe is pretty heavy without the CA adjustment.


What do you think? Seems about right for this type of shot, or do I have something going on here with my setup..? Would shooting faster help? Higher ISO, f/5.6 instead of f/11?


Well thanks for taking a look. Any thoughts you'd care to share - I would really appreciate it.


[img]/cfs-file.ashx/__key/CommunityServer.Components.UserFiles/00.00.00.29.05/_5F00_MG_5F00_8243_2D00_44_5F00_qrtrsz.jpg[/img]


[img]/cfs-file.ashx/__key/CommunityServer.Components.UserFiles/00.00.00.29.05/_5F00_MG_5F00_8243_2D00_44_5F00_100_2500_crop_5F00 _no_5F00_CA_5F00_adj.jpg[/img]


[img]/cfs-file.ashx/__key/CommunityServer.Components.UserFiles/00.00.00.29.05/_5F00_MG_5F00_8243_2D00_44_5F00_100_2500_crop_5F00 _CA_5F00_adj_5F00_rc_5F002D00_63.jpg[/img]

[canoli]

by the way, I am aware this question falls under the heading of "pixel-peeping" in the extreme...

[Benjamin]

I'm not sure if I will get the same effect with my 16-35L II in the same condition. But according to my experience if I see something like this I won't be worried. The 17-40L, along with the 16-35L both have some color fringes, so there is no surprise here. The 24-70L as I have noticed will produce even more red fringing especially when wide open.


I would say that your lens is just fine. Though haven't seen anything exactly like this since I barely took anything that exactly contains the object you have in here, but the slight red fringe you have here does not look abnormal at all.

[canoli]

Thanks Benjamin - I really appreciate your reply. I'm beginning to feel ridiculous for asking.


The more I look at the entire scene...if there's fringing around those distant clouds, I sure can't see it unless I'm viewing pixels at 100%.


and at 100% on a 24" display... all kinds of "horrendous" defects are visible...


I guess if I wanted to produce a museum-quality, mega-sized print, I would've left the 40D at home and shot medium or large format.


(or just Photoshop the 1-pixel fringe out...!)

[Colin]

don't feel ridiculous for being critical. I sent my 24-70mm back for out of focus CA issues. I dumped my 50mm f/1.2 because it was working 'correctly', in that it habitually back focused at aperture between 1.2 and 5.6. I sent back a copy of a 135 f/2 because it backfocused, even though Canon told me I should just stop down.


I mean, I can accept limitations if thats the best that they're going to be, though I see no reason to settle for equipment working less than the best that it can. No lens is perfect, but don't feel silly about being critical.


If you had just a 'little' dust on your sensor, would you feel ridiculous about removing it? Or, if you had hot pixels, wouldn't you want to map them out?


Benjamin has a good point, to a point, but don't settle if you don't have to.

[canoli]

Thanks Colin - I appreciate your support. I'm trying to decide if I should "settle."


I was surprised to see such an obvious red fringe along the treeline and in the clouds, but as it's only visible at 100% on a computer monitor, I think maybe I've fallen into the pixel-peeping trap of being overly critical...


...especially of an image like this, where the detail I'm "expecting" to hold is either in dark shadow, surrounded by sunset colors, very far away...or all three.


So is this normal and expected - to see a 1 or 1-1/2 pixel red fringe in a shot like this, or am I being silly?

[Benjamin]

I mean, I can accept limitations if thats the best that they're going to be, though I see no reason to settle for equipment working less than the best that it can.
<div style="clear: both;"]</div>

I think Colin got it exactly right here.


I guess no one would like to have an underperforming lens at the first place; but most people won't bother to deal with the hassle of sending a lens back, wait for another copy to arrive and then do all the tests necessary to determine if the new copy is a good copy or not...


If I'm not that lazy, all my lenses and cameras should have gone back to the factory at least once... On the other hand, however, I don't see my pictures are compromised because of this, maybe I have a relatively loose standard.


That's just me...

[Colin]

I really don't know if it's normal.


It doesn't seem like awful performance or anything. That is an awful lot of contrast to deal with. An error of only a pixel or two might be really fantastic performance.


I wish i could be more helpful.

[canoli]

You have been very helpful. By your (and Benj) answer I've come to the conclusion that I should probably just rely on common sense.


Like you said, that is an "awful lot of contrast to deal with."


I figured a treeline in front of a sunset glow ought to show any CA / fringing if it was there. And it turns out it is.


But in not-so-extreme shots I've been very happy with this lens' performance. Maybe it's time to quit testing and just make pictures!


Thanks again you guys.