Problem with 24-70 F2.8L USM

[Mikeinminn]

Hi all. New to the forum. I just bought a used 24-70 F2.8L USM and noticed that even the center sharpness at 24-35mm F2.8 was poor.

Obviously, the weakest spot in a lens is at the wide end but the difference between pictures is stark and surprising in light of the ISO chart for this lens.

Photos of the wheel were taken in shaded outdoor morning light manually focused on Live View at F2.8 24mm 1/250 (zoomed all the way in on the camera's review screen) and about 40mm F2.8 1/250 and backed away so that the object takes up roughly the same amount of sensor space. The problem only shows up at when the lens is opened up and at infinity. It looks and feels like it hits the focus limiter before it comes to focus from 24-35mm or so.

This may be happening at infinity at longer focal lengths too as the pictures of the tree show. The tree photos are all taken at 50mm infinity with different F ratios starting with F2.8 then F4 then F8. All pics are approx. 100% crops which are then re-sized and compressed for use here. When I stop the lens down, performance between focal lengths is similar.

Is this sort of performance normal for this lens? Is the wider aperture somehow showing up a focus problem that narrower apertures mask? I'm just trying to decide if this should be returned or just sent to Canon. If Canon charges too much, I may as well just pony up for any new ones that are still available considering the price that is common for a used one. Thanks for any input.

[Mikeinminn]

By the way, all of the pics were taken from a tripod and are crops taken from close to the center of the image.

[Mikeinminn]

Anyone have a guess on this?

[Daniel Browning]

Is this sort of performance normal for this lens?

Definitely not. It looks very out of focus, even for a 40D.

It looks and feels like it hits the focus limiter before it comes to focus from 24-35mm or so.

If it was, that would definitely explain why it looks so out of focus.

Is the wider aperture somehow showing up a focus problem that narrower apertures mask?

Yes, the deeper DOF of narrow apertures can fix defocus blur.

I'm just trying to decide if this should be returned or just sent to Canon.

I would definitely try to return it first.

[Mikeinminn]

Thank you very much, Daniel.

[Sean Setters]

I'll also add that you don't want to test the sharpness of a lens at an ISO higher than 100 or 200 ISO. What ISO were you using when shooting these images?

[Mikeinminn]

I just checked my shot settings. It was set at 800 ISO which was probably a holdover from the night before. Oops. Rookie mistake from a rookie

Still, the comparison should still be valid since I was just comparing one aperture to another on the same lens. The defocus at infinity was very apparent even on the live view screen, not just in the photos.

Or am I wrong?

Thanks for the input.

[Sean Setters]

If you take a picture of an object at 70mm, then walk toward the object letting it fill about the same amount of area within the viewfinder, how are you focusing at or near infinity?

[Mikeinminn]

Actually with the first two (the wheel), I started at 24mm infinity then walked away until the object filled approximately the same amt of the sensor at 40mm. It was not quite at infinity at 40mm, thus the room for accurate focus. The tree was shot all at infinity from the same location at 50mm with F2.8, F4, and F8. The tire was the only one that I moved the tripod for, and only the badly defocused shot was at infinity F2.8. I presented the two types of shot to provide clues as to what might be happening. Perhaps I confused the issue more than I clarified it.

At first, I had thought that the issue was just a focal length+aperture combination issue that might show up only at 24mm infinity. Then I took the 3 shots of the tree and found that even at 50mm infinity the same thing was happening at F2.8 especially. So from this, I gather that it is hitting the focus limiter before it reaches focus. So it's going back to KEH for repair.

[Sean Setters]

I'm still baffled by your testing. Why are you testing a lens's focusing ability at infinity? Before you send it back, do this:

1) Tape a piece of paper with black marker writing on it (big letters) onto a wall. Put as much light on it as you can.

2) Set up a tripod about 10 feet away. To test autofocus, use the center point to focus on the paper. Disengage autofocus and turn the autofocus ring so that it isn't locked onto the paper. Zoom in, turn autofocus on, and focus on the paper again. Analyze the pictures to check for focus accuracy. (Really, you shold do this step several times to get a decent sample size).

3) To check for softness, do the same thing except use Live View x10 to manually focus on the target. Once focus has been achieved, take the picture.

4) Analyze results.