[BigMackCam]

I'm hoping one of you good folks can educate me before I send a lens back for replacement or warranty service.

Just after Christmas, I bought a Tamron 28-75 f/2.8 for my A-mount gear, as I've been very happy with that same lens in K-mount.

However, while doing AF accuracy tests I've found a problem. See attached pic... When focused where that red circle is, the field of acceptable focus (if that makes sense) changes across the frame from left to right. On the left side of the shot, you can see the the text in front of the focus point is in focus, while on the right side, the text behind the focus point is in focus. This is very visible with focus peaking in the EVF, without even having to take the shot.

I don't see this with my K-mount copy running the same test with the camera mounted in the same position, nor do I see it with my other A-mount lenses. I can repeat the test with the same result every time.

Is this a de-centering issue, or something else?

Attached Images

 

---

Last edited by BigMackCam; 01-11-2017 at 02:43 PM.

[MSL]

No answers, but in order to aid those who might be able to answer - what were the settings used for this shot, and does this effect occur at different focal lengths?

[BigMackCam]

No answers, but in order to aid those who might be able to answer - what were the settings used for this shot, and does this effect occur at different focal lengths?

Apologies, I should have given that info. This was at 75mm, f/2.8. It happens at other focal lengths and apertures too, just with greater overall depth of field as you'd expect...

[photoptimist]

Are you sure everything was square, flat, level, and aligned? It's strange that the vanishing point for the two columns of text seems to point to the right.

[rrstuff]

I am afraid this is likely a tilted element, which I guess is kind of decentering. Another kind is when the best focus is off the center axis of the lens. I would take a look if there are any signs of the lens falling on the axis of the tilt in the focus field. Another option could be loose cams or maybe busted helicoid. Either way, in my opinion the lens is mechanically damaged.

[caliscouser]

Are you sure everything was square, flat, level, and aligned? It's strange that the vanishing point for the two columns of text seems to point to the right.

+1

Chart looks crooked to me.

[BigMackCam]

Are you sure everything was square, flat, level, and aligned? It's strange that the vanishing point for the two columns of text seems to point to the right.

Not 100% straight, admittedly (my tests were hand-held), but straight enough that I do not get the problem with other lenses in the same test, and the results I'm getting are repeatable.

[rawr]

de-centering issue

If it was de-centering, you should see the problem on a totally flat test target too.

[Ian Stuart Forsyth]

You have it in tilt, first look for the knob on the lens and unlock it and center the lens back to 0
Just kidding, I would look to see if the lens mount is sitting flush with the camera mount or look to see if the lens mount is flat to the lens

[photoptimist]

Not 100% straight, admittedly (my tests were hand-held), but straight enough that I do not get the problem with other lenses in the same test, and the results I'm getting are repeatable.

Hmmm... That is suspicious. How about at ∞? If you point the camera at a distant horizon, live-view zoom the far right edge and manually focus, then move the zoomed view to the left side, is the left side also in focus or do you have to refocus to get that sharp?