[coolmast]

Hey folks,

during my last holidays I used my new Sigma 10-20 f4.0-5.6 a lot and really got some great shots. Only while processing me from time to time I was heavily disappointed by the lack of sharpness. Until now I thought this would be a mixture of misaligned focus, motion blur and being the lens wideangled a design problem.
However, I rethought all that again and had a look on one of my very first shots with this lens. What you see below is an out of camera jpeg which has (at least for my eyes) heavily more unsharpness on the left half of the image than on the right letting me think about decentering issues.
As I am not an expert on optical lens quailty I wanted to hear your opinions of what was wrong in this shot or what is wrong with the lens?


taken at F/7.1 1/250 10mm ISO200 - Pentax K3

Thanks in advance,
Peter

[othar]

I think you have to post a crop of the left and the right side of the picture for comparison, at this size I hardly see any difference.

Regards, Patrick

[coolmast]

You are right I could have come up with that on my own ;-).
However, I made a crop of the far left and right sides of the image at equal height. The distance to the border of the image is the same for both crops. See below.

[enoeske]

I wouldn't accept that lens. Look pretty badly decentered.

[othar]

I am no expert either but I think your are right, looks like decentering.

[coolmast]

I was not intending to replace it. Instead, since I am still inside the warranty period I wanted to have it checked by Sigma if it's worth it.

[realitarian]

I was not intending to replace it. Instead, since I am still inside the warranty period I wanted to have it checked by Sigma if it's worth it.

Just to eliminate another possibility: was your camera mounted on a tripod or otherwise well supported, and if so, was Shake Reduction on? I recently had a bunch of pictures come out bad because of this - in each of the images, part of the picture was nice and sharp and the other part looked like it was something in between motion blur and out of focus.

[coolmast]

The shot was definitely handheld and if I am not totally off Shake Reduction was enabled.

[JimmyDranox]

First, the advice to take some shots of a Zeiss Siemens chart, or another chart is a very good one. But if you don't have a chart, you can try to shoot the sky. it's easy to see decentration on the stars, even if you cannot focus very well.

Second, even if most of the time decentration of the lens is the most common cause, sometime, the bayonet from the camera can be decentered (tilted), when not all the screws are tightened properly, especially on WR cameras where is a rubber ring under the bayonet. It has happened to me on my former K-5IIs. On medium focal length is not easy to see, but on very wide, this defect can be easily seen. Also, on long lens in special condition. If it this the case, it's somehow easy to solve at home, if you had a chart, or with stars above.

[Digitalis]

I think you have to post a crop of the left and the right side of the picture for comparison, at this size I hardly see any difference.

I could see the issue straight away...get that lens replaced. It is apparent that there is some degree of optical misalignment causing the blur in the upper left corner, at f/7.1 DOF should be adequate for that subject distance, at and 1/250th the chance for camera shake affecting the image is significantly reduced.

if you don't have a chart, you can try to shoot the sky. it's easy to see decentration on the stars

This is a really bad idea.

[tibbitts]

Having this same lens, and a similar experience, I can tell you that Sigma will test it the same way you did: go outside and take some pictures. They'll also photograph a resolution chart - but the test won't venture far toward the edges/corners. Don't be surprised if the conclusion is that the lens is "within specs." Warranties are great when the front barrel of your lens falls off (unless perhaps it appears to have been abused); they're not so great when you're unhappy with the performance in one part of the image or another.

Before you send your lens in, take many more pictures, particularly at different focal lengths. One test I did with mine was to photograph a distant traffic sign and move the camera (it was tripod-mounted) so that the same sign was in each corner of the image. MF, no SR. It's possible that you'll get results like you showed here sometimes, and then other times you'll get better results - amazingly sometimes at identical focal lengths. I can't explain that, but I'm just saying don't be surprised if it happens. It's very annoying, because if anything, you want a lens to be consistent. But I've had the same experience with other lenses, including Pentax lenses. I think I'm fairly meticulous in testing, yet I can't always precisely reproduce poor results. What you're hoping(?) is to get reasonably consistent results that demonstrate the problem. If you can't do that, it's really not worthwhile sending the lens in.

[osv]

I could see the issue straight away...get that lens replaced. It is apparent that there is some degree of optical misalignment causing the blur in the upper left corner, at f/7.1 DOF should be adequate for that subjectdistance, at and 1/250th the chance for camera shake affecting the image is significantly reduced.

^^^this

shooting a chart means that you must have the sensor perfectly square to the chart, which is a pain in the neck... even outside with wind at 1/250th, both sides should look similar, but it's usually best to shoot it with a fast shutter.

o.p. did a decent job with the subject matter, but next time post a full-size image, not something that has been edited... lenses can have out of focus areas, and wavy field curvature problems, that don't show up with partial crop posts.

i would return this lens for refund or replacement, if at all possible, don't hassle with sending it to the manufacturer unless there is no other choice.