[GregK8]

So today I sat down and tried to do a real comparison to see how much better the DA* 16-50mm was than the kit lens that came with my K20D, and boy was I surprised. Set to 18mm at f4.0 the DA* clearly beats the pants off the kit lens. The difference is astounding, but the surprise is that at a smaller aperture, like f8 or f16 the kit lens comes pretty damn close to the DA* in terms of sharpness, in fact, in my opinion, the 18-55 was noticeably sharper than the 16-50 at 28mm.

It doesn't make sense, so I did some investigating and I have preliminarily determined that the 16-50mm may be back-focusing. So I'll withhold judgement until I can be certain. Is it possible for a lens to back-focus at one focal length but not at another?

The other interesting phenomena I noticed was that with the kit lens, the camera underexposed by between 2/3 and 1 EV. The camera also always selected a warmer white balance with the DA 18-55mm, I believe almost always by 300 Deg. K. Does anyone have some knowledge that could explain that?

For this test I set up on a tripod, used AF and AE for every shot.

[kytra]

No surprises there for me, the pentax kit lenses are pretty darn good (with a few exceptions) and the differences between the DA and the DA* are apparent only when wide open and at smallest and longest FL.

The amount of vignetting and distortion, as well as CAs and flare are surely better controlled in the more expensive and optically superior DA*s but in sheer resolution terms, at small apertures, they could be very close.

If the DA* is back-focusing, that's another story. I think that usually the back or front focusing is constant across the FL, and can by corrected in-camera, to a certain amount.

[GregK8]

There are a lot more advantages to the DA* over the kit lens, most notably:

Weather resistant (This is compared to the non WR version that came with my K20D)
f2.8 throughout the entire FL range
AF is much quieter

and the most important: Size matters

[Marc Sabatella]

It doesn't make sense, so I did some investigating and I have preliminarily determined that the 16-50mm may be back-focusing.

While that's certainly *possible*, no way is that an issue at f/16, or even f/8.

The other interesting phenomena I noticed was that with the kit lens, the camera underexposed by between 2/3 and 1 EV.

Underexposed by what standard? Exposures should be according to ISO standards - about half a stop left of center on histogram on average. Usually when people complain of exposure, it's just a matter of them subjectively preferring something brighter than standard. But you shouldn't be seeing anything close to a 1EV difference in metering on the same scene between lenses. Can you post samples?

[GregK8]

While that's certainly *possible*, no way is that an issue at f/16, or even f/8.

You are correct about this. I probably should have been more clear that the images at 28mm/f4 were the ones that made me question this.

Underexposed by what standard?

One lens compared to the other. I had to boost the EV in PP on the images from the 18-55 to get the same brightness as the ones from the 16-50. I think I can get an example up pretty quickly.

[GregK8]

Here is a comparison at 18mm and f4.0. The first image is the DA* 16-50, the second from the same spot using the 18-55mm kit lens. By my eye the second image needs to have it's EV boosted by +1/3 to equal the first.

[GregK8]

This second set of shots is at 50mm and f4. The image from the 18-55mm needs at least a full +1 of EV to make them the same on my monitor.

Attached Images

   

[GregK8]

So all these images are done with AF and AE, letting the camera do the work. I applied no PP to any of these, I just converted the PEF files with LR3 using default settings for all 4 of them.

[Marc Sabatella]

In the first comparison, while similar, that's not exactly the same scene. Even the different picture displayed on the monitor may have been enough to explain the 1/3 EV difference in metering, but there is also a different in how much background is present.

In the second, most of the difference is presumably because you are shooting a variable maximum aperture zoom at its maximum aperture, and these are always just approximations. As you get to the long end of a given stated maximum aperture, it because more and more optimistic. Note the 18-55 should not even be letting you set f/4 at 50mm - that must really be f/4.5. So there would be half a stop right there - but it's also at the extreme end of the f/4.5 rsnge, and is indeed going to be about as close to f/5.6 as it is to f/4.5

[bdery]

At f8 or f16, any lens will perform well enough. The "cheaper" lenses usually are not really optimized in regards to aperture, which means they will naturally improve as you close down the iris, until diffraction hits (a little before f22, typically). Higherr-end, faster glass will often be optimized at larger apertures, which might lead to some compromises as you close things down.

Different lenses are also optimized for different focal lengths. I'll bet at 50 mm the DA* trumps the kit at any aperture.

[Pentaxor]

honestly, the DA16-50 is still better from the compared images presented. would be interested to see how those two would look at f8 and f16.

[tempelorg]

I recently used a B&W Siemens test pattern to check the sharpness (and effects of diffraction) of my various lens including 18-55 AL II, 18-55 WR, 18-135 WR and DA* 16-50. Even at the center, at same apertures, the (newer) WR models were much sharper than the AL II. And the DA* being still better. In fact, the AL II's quality was quite crappy - even my cheap Tamron and Sigma 18-200 zooms did do better. But I remember reading reviews claiming that the AL II was quite good and maybe better than the WR model. But then, I only tested one case, and maybe it shines in other respects.
Anyhow, just wanted to babble a bit

[Marc Sabatella]

The AL II and the WR are the exact same lens optically; any difference you see is either testing error or just sample variation.

[mikeSF]

hey Greg, is that a Ted Crane pen and ink drawing?? Looks similar in style to one i have, that's all.

[tempelorg]

The AL II and the WR are the exact same lens optically; any difference you see is either testing error or just sample variation.

Wow, does else everyone agree on this? I see significant quality differences between the two of mine. Not that it really matters as I don't have use for the AL II any more, but since I wondered if I should sell it on eBay, I should probably point out the possibility of having a really badly produced individual sample.