24-70mm f/2.8 L focus issues

[tom_bone]

<p>Hello Everyone,<br>

I am new to photo.net though I have been visiting the site for a long time.<br>

I am posting because I am having issues with my Canon 24-70mm f/2.8 L lens I purchased a few months ago. Until recently, I have been using it with a C100 with no problems. Yesterday I attached it to my 550D and experienced focus issues.<br>

Perched on a solid tripod, zoomed to 70mm, while focusing manually on a particular object in live view, the LCD looks tack sharp, but the object in the photo is blurry. However, I can see that an object behind my focus point is tack sharp.<br>

In other words, the actual plane of focus was further than what I has focused manually on.<br>

To test this, I tried focusing in front of my object (so that it looked blurry in live view) and the photo turned out sharp.<br>

Frankly, I am stumped. I thought live view was what the sensor was actually seeing. Can anyone explain to me what is actually happening? and what I can do about it?<br>

I don't seem to be having this problem with the C100, but I am not 100% certain.<br>

Thank you very much for any assistance.<br>

Tom.</p>

[PuppyDigs]

<p>So manual focus via eyesight is off? How about if you use the magnify mode on the LCD? Unless DOF is great, I have to use magnify mode to nail focus manually.</p>

<p>Does it autofocus correctly? I had to MA several of my lenses to get them tack sharp, e.g., my 70-200 4L IS required +3 at 70mm and +2 at 200. If you camera lacks the MA feature, you'll have to send the camera and lens to Canon for calibration.</p>

[Robin Smith]

<p>Is this a shutter speed issue? What is the object you are focusing on?</p>

[tom_bone]

<p>I am manually focusing on (relatively close) stationary objects with 10x magnify mode on the LCD. I do not have MA on my camera.<br>

Upon further testing, I have noticed that at 70mm f2.8 the image is sharp; at f6.3 (and higher) the image is off (back-focusing). So something is happening in between.<br>

The only way I can explain this to myself is that my lens suffers from focus shift as I step it down from 2.8 to 6.3 and that the Live View LCD image is always 2.8 - regardless if the camera is set to something smaller.</p>

[tom_bone]

<p>Robin: No, it is not a shutter speed issue, as I mentioned, the objects behind my focal point are sharp. The lens is clearly back-focusing. I was under the impression that back focusing is only an auto-focus issue.<br /> Given the fact that I am manually focusing and that there is a different in focal plane from 2.8 to 6.3, I think this is a question of focus shift that I am not picking up on with live view.</p>

[peter_langfelder]

<p>Tom, I think you nailed it. There may be significant focus shift when stopping down (also known as residual spherical aberation). Trying focusing at the working aperture, not wide open (or shoot wide open when focusing wide open). The focus shift due to RSA often seems to be most pronounced at short focusing distances. <br>

<br />I am familiar with the 550D but on my cameras (5D MII, 70D) in live view the lens is stopped down to the aperture you set on the camera, not wide open.</p>

[tom_bone]

<p>Is there a known cure for residual spherical aberration?</p>

<p>If I send it in to Canon, would they be able to adjust it or is it something that is fundamentally flawed in the design of the lens or in the particular components that make up mine?</p>

[Robin Smith]

<p>What's the distance to the object? Frankly I am amazed that any focus shift would make such a huge difference unless you are really close? I also don't know of any "focus shift issues" with the 24-70mm particularly. Sounds like there's something wrong somewhere.</p>

[tom_bone]

<p>The distance to the object was between 18" and 24" (I have since taken down my set up so the distance is approximate). I can tell you that the focus shift was large enough to render it a useless image. Godd thing I had my 100mm with me to do the job.<br>

The other thing to mention is that the shift was larger the narrower the field of view, so 70mm was more pronounced than 50mm, and so on.<br>

I am still using the lens day to day for some video work so I cannot send it in to Canon for servicing until next month. I am hoping this is something they will be able to adjust.<br>

If anyone has ever sent in their lens for "residual spherical aberration" repair I would be interested to hear about your experience.<br>

Tom.</p>

<p> </p>

[Robin Smith]

<p>Tom</p>

<p>Spherical aberration means the edges will not be able to be focused sharply at all (the center may be). Which doesn't sound like is the case here. Anyway, there's no cure unless the lens is otherwise damaged and not up to spec. There may indeed be some focus shift at very close distances, but if this is version 1 of the 24-70 I have not noticed it.</p>

[tom_bone]

<p>Robin: I may be mistaken, but although wide open (2.8) allows light from the edges of the lens to hit the focal plane, it has little to do with how the edge and the centre of the image look.</p>

<p>What it happening with lens shift is that the light entering from the edge of the lens converge at a different point than those entering the centre of the lens, this means that the point of focus (circle of confusion) will appear differently than when stepped down (when the edge rays have been eliminated). Basically, the point that looks sharp at 2.8 is different from the point thaat looks sharp at 6.3. The end result is that when you focus at 2.8 you end up focusing in front of the point that would have looked in focus at 6.3.</p>

<p>This is difficult to describe without a diagram, I have found a useful one here: http://photographylife.com/what-is-focus-shift <br>

By the way, my lens is version II.</p>

[Marcus Ian]

<p>It depends on your LV settings, but if you set your LV function to stills + movie, it should stop down the lens as you change the aperture, this may give you an idea whether a focus shift is your problem. With this lens, if you are experiencing focus shift, this is a problem with your copy of the lens.</p>

<p>Alternatively, if you simply use your DOF preview button in LV, you should be able to see whether the point of focus is shifting. </p>

<p>The spherical aberation is a 'problem' with this lens, but typically only WO and near. @ f6.3 it shouldn't be even remotely detectable, much less cause this much of an issue.</p>

[tom_bone]

<p>Marcus: Thank you for your comments. Unfortunately, there is no "stills + movie" function in LV for the T2i so there is no real way around the lens shift that I can think of with my camera. </p>

<p>I also thought of using the DOF preview button - as you suggested - but I didn't notice the lens shift. I was surprised, I thought I would.</p>

<p>As I mentioned, I will send my lens to Canon for analysis and, hopefully, repair. It is too bad about the lens shift, it is otherwise a very beautiful lens to work with.</p>

[philip_wilson]

<p>Tom I have had this lens for many years and still like it. It does suffer from some field curvature however. While I have never shot mine that close it may be part of the issue - although why it happens on one body and not the other is perplexing. While my copy does not seem as bad as the one Photozone tested they do a good job of documenting what they found. Could this be part of your issue? http://www.photozone.de/canon_eos_ff/528-canon2470f28ff?start=1</p>

<p> </p>