Testing Wide Angle Lenses

[mtngal]

Can you tell the side to side sharpness of a wide angle lens by shooting brick walls or the equivalent? I just got a replacement DA*16-50 lens (I sent the first one back to where I bought it) and while it's significantly better than the first one, I can't believe that the pictures I'm seeing are right. The first lens I had wasn't sharp anywhere, and more so on the left side. This lens is really sharp in the center (I can see why people who have good ones love this lens), but is a bit blurry on the left and very blurry on the right. I'm assuming that I have another decentered lens. That got me wondering if walls are really a valid test - the sides are further away than the center. However, in this case, I also took some pictures of a ridge that's a good mile away - I saw the same blurriness on the sides and it should all be in focus - it's well into the infinity range, no dof changes. But it did make me wonder about whether brick walls are truly a good test.

Can this type of thing be caused by damage during shipment? While the inner box was fine, the outer box looked like it had been dropped pretty hard - the side around one end had 3 crinkles - pretty significant ones.

I think I'll ask B&H if I can get a refund instead of another exchange. One of the reasons I wanted to buy this lens right now is that I can get the K-7 rebate. Unfortunately, I bought two lenses that qualify and you have to send everything in within 30 days of buying the lenses (everything includes the bar code for the lenses) and I don't B&H will ship me another replacement lens in time to have the other lens I bought qualify. I can always use the money for something else and buy this lens later. Of the 4 lenses I ordered, this was the one that I felt the biggest need for (I often miss having something between 24 and 50mm) and the one that excited me the least.

I'm really disappointed as I had high hopes this second one would be a good copy and I'd be delighted with it. The only other option would be to keep the lens, send in for the rebate and at the same time send the lens into Pentax for repair. I'm just not so sure that's a great answer - keeping a lens that's faulty right out of the box - think I'd rather get the rebate. Sigh.

Does anyone have any suggestions, thoughts or words of encouragement?

[stevebrot]

Brick walls are useful, but not a real good test. Newsprint (say the classified section of the evening paper) at 20x the focal length is better. Make sure the setup is well-constrained (tripod, 2 second delay, shutter speed above 1/125s) with the camera parallel to the target. You can expect some edge softness with most lenses. What you should not see is asymmetry in the region of sharpness. This would indicate the centering issues that have plagued the 16-50.

Steve

[mtngal]

Thanks for the suggestion about using newsprint and the distance - I'll try that tomorrow morning before I make a final decision.

[Ash]

Sorry to hear this.
If you're already noticing symptoms of decentering on casual shooting, you'll probably exemplify them on a more formal test.
Do show us how you go with the test.

[mtngal]

Interesting shooting the newspaper. It might not have been completely square, but it was as close as I could get by eye without having a T-square handy and something to measure the axis of the camera's centerline. There seemed to be far less difference between the right and left side than what I had noticed with the brick wall pictures yesterday. That means I'm even more on the fence about this lens than I was. Final decision needs to wait until I get home this evening and can look at them on my monitor at home though.

I'd like to post a couple of them and ask whether the fall-off in sharpness is typical for this lens - I could be asking too much of it. The edges are significantly worse than the center, and I would have thought that at this price range, it shouldn't be that much. And I discovered that having SR on when the camera is mounted on a tripod does make a difference - I forgot to turn it off at first and those pictures were not consistent (found that interesting). They were consistent once I turned it off. Also, I was using a flash for extra lighting and wonder if the light drop-off at 16mm also would play a part with sharpness - it made a difference as far as CA (the edges have more CA than I've come to expect with most of Pentax's newer lenses, but I've certainly seen worse).

If anyone else has some ideas for my lunch-time shooting, I'd appreciate it. I need all the help I can get!

[opiedog]

could you please fill us in with your lens' serial number?
i think there is a link somewhere on the forums that tracks the 16-50's serial numbers goo/bad etc.

[mtngal]

I'm still not entirely sure about this lens. The center is really sharp, focus seems very good but the edges still bother me. I'd love to have some other opinions about some test shots along with some walk-around type shots.

These are all jpg straight out of the camera, no processing at all.

First, a brick wall (36mm, f5.6, 1//1000 sec, taken from about 10 - 20 feet away handheld):



full sized picture at: http://mtngal.zenfolio.com/img/v4/p1058143171.jpg

Newspaper - tripod mounted, SR off, remote. Newspaper was taped onto a vertical surface, tripod/camera as close to square to the surface as I could tell. 16mm, f4, 1/50 sec. flash:



full sized at: http://mtngal.zenfolio.com/img/v3/p891465709.jpg

31mm, f4, 1/80 sec, flash:



full sized at: http://mtngal.zenfolio.com/img/v4/p878244330.jpg

50mm, f5.6, 1/100 sec. flash:



Full sized at: http://mtngal.zenfolio.com/img/v0/p884508295.jpg

Some walk-about shots:



full sized: http://mtngal.zenfolio.com/img/v0/p810237425.jpg



It's worth it to look at this one full sized because you can actually see the surface texture of the wall behind the lettering (something I found impressive)

full sized: http://mtngal.zenfolio.com/img/v4/p929414491-4.jpg



full sized: http://mtngal.zenfolio.com/img/v1/p545134998.jpg

These weren't the worst, some are the best. The center of the lens is really good, it's only the sides that concern me. On first glance I really liked the theater sign, but if you look at the full sized picture, the right side is a bit blurrier than the left.

If I were to only use the lens in the center, I'd keep it in a heart-beat. I thought this picture was as good as it gets:



But I do take lots of landscape and some architecture and so full frame sharpness is pretty important to me. Are these walk-about pictures typical for this lens? Am I being silly to want more than what I have here? Should I send the lens back and buy another one in a year or so, or keep this one and hope it doesn't get worse? I'm really on the fence with this, so any help/suggestions/opinions would be greatly appreciated.