How to check your new wide angle for decentering?

[derek_thornton1]

<p>I will be getting the Tamron 15-30mm f/2.8 soon. I have read in a couple places to check the lens when you first get it for decentering. Is this lens known for this? If so, why are there so many great reviews on it? Anyway, I saw a video on decentering from Lens Rentals. It was somewhat informative but the little test sheet he had would be worthless with a wide angle unless you were a foot or two away.</p>

<p>So I would appreciate any instructions on how to test a wide angle lens for decentering.</p>

<p>Derek</p>

[ShunCheung]

<p>I would put your camera on a tripod and parallel to some brick wall. If you are really picky, you can set up a mirror on the wall and you should be able to see the camera's reflection from the mirror through the viewfinder when the camera is totally parallel to the wall.</p>

<p>Capture some test images at different focal lengths, e.g. 30mm, 24mm, 20mm and especially 15mm. If samples at 15mm and f8 or so are sharper in one corner than another, there could be some decentering issue. At f2.8, I would expect the center is sharper than the corners; that I would consider normal.</p>

[JDMvW]

<p>My mother used to say "Don't borrow trouble". That is, respond to trouble when it <em>actually</em> occurs rather than looking for it.<br>

Most lenses are pretty good right out of the box, unless there has been some sort of damage. I'd go out and shoot some images and examine those before panicking.<br>

At least initially, Shun's suggestion should show any potential problems.</p>

[rodeo_joe1]

I have to disagree with JDM here slightly. I think it's

better to thoroughly test a lens as soon as you get

it, rather than find a problem down the line when it's too late to return it. Those test shots can also act as a reference if, say, the lens is dropped or goes out of alignment in some other way.

 

A fairly easy way to check for decentring is to set up the camera on a tripod vertically in manual focus. Point the camera at any distant detailed subject that fills the frame into the corners (doesn't have to be a flat subject). Use a wide aperture and focus accurately on an easily recognised spot. Take a shot and then, being careful not to nudge the focus, swing the camera through 180 degrees and take another (now upside down) shot.

 

When you examine the two shots the same areas of the subject should match in sharpness. The corner and side to side sharpness especially should be identical on both.

 

You can repeat this procedure using various focus points on the subject just to be sure.

 

The brick wall test works as well, but it's often difficult to find a big enough expanse of wall to properly test a wide angle. There might also be some lack of parallelism and field curvature that confuses the issue.

 

I wouldn't worry too much about a tiny amount of decentring that can only be seen at 100% pixel peeping. Most lenses show a small degree of decentring, especially zooms and aspherics. Noticeable decentring can make bokeh look ugly though, and affect depth of field.

[ilkka_nissila]

<p>I believe decentering is easiest to detect at long distances (near infinity focus). I have only had one example of a annoyingly decentered lens (one where the effect would bother me in everyday applications) but that lens was originally designed for 35mm film only showed uneven sharpness on digital, not on 35mm film. </p>

<p>I've noticed a slight tendency for a tilted plane of focus in another lens when photographing the Milky Way, but that I believe to be a question of how precisely the elements are held in position at different angles (relative to gravity) and the effect doesn't show in my normal photography, only in this specific application.</p>

<p>I'm unfortunately not familiar with the Tamron. I guess the important thing is to find out whether your lens has such decentering which would bother you in real world applications which you are likely to do. Some long distance landscape shots should reveal whether your lens is fine. I would also do some testing in your normal photography applications, just shoot normally and see how the results look in general. It's important to be aware that there may be some field curvature and you should not expect that the sharpness is optimal in a plane but it can be a curved surface of some shape. If the sharpness appears highly asymmetrical then it can be that you have a problem lens that needs to be replaced or repaired.</p>

[dshombert]

<p>I haven't tried this, but I've read that a good test for decentering is a shot of some foliage - trees with a lot of leaves all pretty much the same size. Fill the frame with them. If it's decentered, it'll be fairly obvious. As I say, I haven't tried it, but it sort of makes sense. I guess.</p>