Help! - 70-200L 2.8 IS Chromatic Aberration?

<p>I just bought a 70-200L 2.6 IS for my 20D and I've noticed what

seems to

be too much chromatic aberration in my shots. Comments from anyone

experienced

with this lens or problem would be much appreciated.<br>

<br>

What I'm seeing is red/blue fringe, even in the center of the image

(4 pixels).

In the lower right corner I'm getting 7-8 pixels of the same. The

cleanest

region is 1/3 up and 1/3 to the left of center, but even it has ~2

pixels of

fringe.<br>

<br>

With all the great reviews of this lens, I'm assuming this is not

normal. I can

see the issue in real-world shots, and my cheapo 55-200 II lens has

less fringe than the $1600 lens does.<br>

<br>

Below are some <a href="http://www.photo.net/photodb/folder?

folder_id=565146">

samples</a> :<br>

<br>

70-200L 2.8 IS lower right and center, respectively (200%):</p>

<p>

<img height="404" src="http://gallery.photo.net/photo/4060929-

lg.jpg" width="398" border="0"> 

<img height="400" src="http://gallery.photo.net/photo/4060922-

lg.jpg" width="400" border="0"></p>

<p>55-200 II lower right and center, respectively (200%):</p>

<p>

<img height="430" src="http://gallery.photo.net/photo/4060941-

lg.jpg" width="408" border="0"> 

<img height="420" src="http://gallery.photo.net/photo/4060939-

lg.jpg" width="434" border="0"></p>

<p>Are you seeing this in your 70-200L 2.8 IS? Any comments are

appreciated.</p>

<p>-Steve</p>

This isn't really a surprise. The 70-200 f/2.8L IS is loaded with many elements.

 

Luckily for you it appears you have a digital camera. You can 100% correct CA if you are

willing to shoot RAW files.

Steve...I have found the same affects with my 75-300 (and my 20D). One of the reasons I recently purchased 2 new L lenses (17-40 and 100-400). I hope the added cost is worth it...I'm told it is. I hear the difference in clarity is quite noticeable...let's hope.

Andrew, I've tried Photoshop CS2's CA correction in the RAW import, and it doesn't get rid of the fringe. It only modifies it - i.e. yielding green/cyan/magenta fringe on different axes while reducing the red/blue. Your "100% correct CA" makes me think I'm missing something. Is there a trick to using the feature?

 

I'm wondering if the Photoshop feature is designed to correct Longitudinal CA (colors focusing in different planes) and my problem is Lateral CA (colors focusing at different targets in the same plane). I suspect the latter because I have asymmetric fringing at the center of the image - that is, all red fringe on the upper left and all blue on the lower right.

 

-Steve

I don't have the 70-200/2.8 L IS (I own the 70-200/4L) but I seriously doubt that this is this normal. The chromatic aberration reported by PhotoZone is less that 1 pixel (on a 350D not a 20D but they are nearly the same in terms of pixel pitch) except with a teleconvertor mounted.

Hmmmmm. Hopefully something is going wrong in the conversion heres my experience handheld quick at 135mm f/2.8 1/30th sec. 135mm f/2.8 having the most CA according to Photozone.de . 400% and 1500% crops were done upsizing in photoshop with Nearest Neighbor Sampling.

 

Hopefully pictures appear below:<div></div>

100% Crop<div></div>

400% Crop<div></div>

1500% Crop<div></div>

Defineately not a controlled test and certainly not as high in contrast as your pics but I'll let you draw your own conclusions.

Could someone suggest me how I can test my lens and especially if there are are any targets I can print to test?

 

Thanx in advance.

The trick with the PSCS2 RAW CA correction is to use ONLY JUST the right amount of

correction needed. No more, no less.

<p>I bought a new 70-200 IS last month and haven't had a lot of opportunities to play with it. I did notice CA, but nowhere near what you're seeing; my 70-200's CA is well controlled, and probably comparable to what you're showing in your 55-200 photos.</p>

Jan, the CA in your shots indeed looks minimal. Note though that typical Longitudinal CA will appear most noticeably on tangential lines, or those that "face" the center of the image. Since the subject in your shot had lines at 45 degrees from that, the CA you'd normally expect is probably about 40% more than you see. But nonetheless your CA still looks great and I'd be delighted if my 70-200 looked that way.

 

I called Canon technical support yesterday and described the issue. They concluded that it was a definite problem with the lens, so I sent it in for warranty repair.

Steve, thanks for the feedback about your own 70-200 IS. I look forward to getting mine back from repair and having it perform to its "L" reputation and cost.

 

-Steve

<p>Here's a sample from mine. This was handheld so it's a bit blurry but clearly there's CA there, as the blur on either side is of a different colour. This is from the very corner (on my 1.6-crop 20D) of the frame, 200mm, f/4, and it's a 100% crop from a large/fine JPEG.</p>

Steve, can I assume that that was the lower right corner? If so, the kind of aberration seem consistent with mine. Although I'd say mine is at least 50% greater in width. I don't know if the added saturation in mine is a function of the contrast or just more aberration and therefore purer reds and blues not diluted by the other frequencies.

 

Is CA virtually undetectable in the center on yours?

<p>Sorry about the delay in responding. The above picture is from the very lower right corner on my 20D. The centre of that image isn't much use for testing CA; most of the image (a scaled-down low-quality version of the whole frame is shown below) consists of a Macbeth-type colour chart, and all of the black/white contrasting areas are well away from the centre.</p>

 

<p>From looking at some other test shots I took (albeit of much lower-contrast subjects), I'd have to say that it doesn't look like there's CA at the centre of the image, which isn't really a surprise.</p><div></div>