0
\$\begingroup\$

I have been having focus issues with a Canon EF 24-70mm f/4L IS USM lens.

It seems that it is unable to focus at infinity (or close to it)for most of the range between the wide and telephoto ends below ƒ8.

I have sent it away to be repaired and had to return it twice because the problem hasn't been fixed. I've received it back for the third time and the problem is still visible. The last time I sent the lens I sent the body too. The lens shows the same issue on different bodies.

If I pick a distant object and use the 10x zoom on the live view to manually focus, the focus hits infinity before the object is fully in focus.

This is my main lens, and all in all it has now spent 2 months away at the repairers whilst they wait for parts. I can't spend another month waiting given that given their track record it will almost certainly still have issues when I get it back. I'm at a loss as to what to do. Do I accept the lens is unfixable and buy another? The lens is less than two years old, so this would be a worst case solution. Am I being unrealistic? Is this just expected diffraction from a lower ƒ number? What else can I do to test the lens?

Here are 2 (Raw) test images manually focused using live view at 10x zoom on the towerblock @ 52mm. Shot on a tripod with a cable release on a day with no wind.

Details from each image:

enter image description here
Image taken at ƒ8 RAW

enter image description here
Image taken at ƒ4 RAW

Improve this question
\$\endgroup\$
3
  • \$\begingroup\$ You say these shots were manually focused. In that case It's your fault if the subject isn't in focus. How is this the lens' fault? \$\endgroup\$
    – Olin Lathrop
    Jan 17, 2016 at 14:21
  • 5
    \$\begingroup\$ @OlinLathrop: The lens/focus ring hits infinity before it is in focus. @ Pedr: There is an option to reduce the flange-to-film distance by modifying the mount of the lens. Not very elegant but it would solve your problem. Probably there also can be some other mecahnical mean to reduce the distance, like tightening some screws. For that however you need to be sure that the unsharp image is not just a result of the resolution limit of the lens. Have you compared this lens to other ones of its kind and their resolution limit at f4? Do images at a less far distance get sharper at f4? \$\endgroup\$
    – kamuro
    Jan 17, 2016 at 14:27
  • \$\begingroup\$ @kamuro Thanks for your comment. It looks like a general softness in the middle of the zoom at ƒ4 that improves at ƒ5.6 and is fine by ƒ8. \$\endgroup\$
    – Undistraction
    Jan 19, 2016 at 17:37

1 Answer 1

Reset to default
3
\$\begingroup\$

The problem has nothing to do with the image not being entirely focus but more of that it is soft. At f/4.0, there is a significant softness that the lens has.
When you are looking at photo 100%, this softness becomes more pronounced. Details just become harder to capture for such a small portion of the image.

Remember that when you are looking through the live view, the aperture will at its widest unless you hit the DOF preview button. That can be used as a quick compare and contrast in the field to see lens softness.

This website has done some extensive testings and will be very useful to see just how much more softer your lens is at f/4 versus f/8 when at 50mm. Canon EF 24-70mm f/4L IS USM 50mm - f/4 versus f/8

If you want to test your lens, search for photography test patterns on the web and print a few different ones out. You can either frame most of the patterns in the frame or try to move back so far that at 100%, only the paper will be in the shot completely, the same as the building was in your shots provided.
Compare the two for sharpness. The f/4 should be softer than the f/8.

The one program that I know that can remotely try to correct for this is DxO. Which, even then, still cannot correct and recover details lost due to lens softness.

Improve this answer
\$\endgroup\$
3
  • \$\begingroup\$ While you are right with all the points you make, none of them can actually rule out if perfect infinity focus is not hit in the TOs case. Let's say he/she is only able to focus to 15m and not beyond, the depth of field stretches to around 50m at F4, but to infinity at F8. While I don't know about the real world characteristics of the lens in question (and wide open, I agree, there will be significant softness due to all kinds aberrations), it still might be possible that the infinity focus condition just cannot be met. \$\endgroup\$
    – kamuro
    Jan 19, 2016 at 23:56
  • \$\begingroup\$ @kamuro I guess one way to see would be to have something large at infinity (say mountain) and see the results. That or see how autofocus does it in live view. \$\endgroup\$
    – Kel
    Jan 20, 2016 at 1:14
  • 1
    \$\begingroup\$ Or the TO could deliberately focus on something which is at less than 20 m at 50mm focal length and see if the same behavior can be found for background and foreground or not. If it is the exact same behavior found for the infinity focused image, it's clear. If not, it might just be softness of the whole image. But also beware of the spot of the image you look at. There are generally more aberrations in the corners of images, than in the middle. \$\endgroup\$
    – kamuro
    Jan 20, 2016 at 8:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.