[stanic]

and another test, this time from germany
Pentax SMC DA 21mm f/3.2 AL Limited - Review / Lab Test

looks like this lens has more field curvature than 15/4, doesn`t it?

[RonHendriks1966]

That the lens was pinsharp in the centre I had already a good nose for:

https://dl.dropbox.com/u/20195454/Son%20-%20optocht/K5D_0112A.jpg

[stanic]

pretty sharp and I like the OOF rendering too

[liukaitc]

Well, not very sharp, based on data....

[philbaum]

To me, this is getting to be an academic exercise, which is not of much real practical value.

Back in the K10 days when most of these Pentax lens tests were first done, the Sigma 10-20 was and still is considered quite sharp. (i have it and like its sharpness). The DA 12-24 was and is considered even better (i don't have it).

Well, with the K5's higher resolution capability with 16 mpix, the current test of DA-21 blows away both of those 2 former tests of the UWA zooms,

I pick and choose lenses to take along based on what i will be doing, except the DA-21 always goes with me, i like it a lot for very detailed imanges across the frame, typically shooting at f8. For street shooting and casual stuff, i really don't care that the borders and extreme edges at f3.2 and f4 are not quite as good as the best of the primes. Life's about compromises and this is one lens that is an excellent compromise.

[asdf]

Well, not very sharp, based on data....

The data is pretty typical for wide angle lenses, though. This is why Zeiss T* 21 is so exceptional and expensive, I guess... In any event, DA 21 is meant to be a compact landscape or cityscape lens. If you're shooting landscape you're shooting at f/8 or smaller apertures.

[stanic]

Well, not very sharp, based on data....

Focus Fallibility: Lens Test Fallacies - SLRgear.com!

I pick and choose lenses to take along based on what i will be doing, except the DA-21 always goes with me, i like it a lot for very detailed imanges across the frame, typically shooting at f8. For street shooting and casual stuff, i really don't care that the borders and extreme edges at f3.2 and f4 are not quite as good as the best of the primes. Life's about compromises and this is one lens that is an excellent compromise.

well said

[fikkser]

In the test done here at pentaxforums I'd say 18-55 wr is sharper than the da 21 when looking at the testcharts. Even when the 18-55 is presumably at biggest aperture and the 21 is stopped down. 18-55 might be worse in extremes at some f, but overall the 21 gets kicked on the nuts by a cheap kit lens in sharpness. The 21 might be slight oof in the test?

[luftfluss]

In the test done here at pentaxforums I'd say 18-55 wr is sharper than the da 21 when looking at the testcharts. Even when the 18-55 is presumably at biggest aperture and the 21 is stopped down. 18-55 might be worse in extremes at some f, but overall the 21 gets kicked on the nuts by a cheap kit lens in sharpness. The 21 might be slight oof in the test?

I see what you're saying. The exposure between the samples is not always consistent, though, and I see some haloing around some of the 18-55's samples, which looks like some PP was done.

There are some good 18-55 lenses out there, unfortunately there are also some less good ones (like the one I have).

[paperbag846]

I'd say the 18-55 is *as* sharp as a lot of lenses *subjectively*, because I think sharpness is one of the most common features of any lens. I've used some real junk-bin lenses that were *sharp* (they wern't anything else, but they were sharp).

I honestly believe that sharpness is the most overrated aspect of lenses (likely a holdover from the yesteryear when sharpness actually was quite variable).

Sure there are a few lenses that are *VERY* sharp, but for the other 90% of lenses, the differences in sharpness do not matter at all.

Now when it comes to colour, DOF control, contrast, flare resistance... there is where you are going to see th difference between the 18-55 and the 21.

[northcoastgreg]

I honestly believe that sharpness is the most overrated aspect of lenses

I find this to be largely true as well. Once a sufficient level of sharpness is reached (and most of the current Pentax lenses reach this level, at least across most of the frame and at most of their focal lengths), additional sharpness is not going to have a great effect on how most images turn out. It's the microcontrast, the color rendition, the overall beauty of the rendering that will make the biggest impact.

One can find plenty examples of lenses with less resolution producing better images than lenses with more. At least in terms of center resolution, the 18-55 is sharper than the DA 10-17, but the 10-17 will tend to produce better images, because it has more contrast and better color rendition. I've seen resolution tests where the old M 50/2 actually outresolved (by a small amount) the M 50/1.7. But the 50/1.7 is clearly the better lenses. The M 50/2, despite its commendable sharpness, produces (relative to the 50/1.7) flat colors and uninspiring contrast.

Lens that have excellent microcontrast and produce rich or striking color often seem sharper, because they produce images with more "snap" or "pop." So I suspect that some people confuse measurable resolution with contrast and color rendition. To judge a lens, one has to look at the images it produces, rather than how well it resolves MTF charts. (How many people buy a lens to shoo MTF charts?) It's the quality the images a lens produces, not how well it scores on numerical tests, that is important.

[stanic]

Now when it comes to colour, DOF control, contrast, flare resistance... there is where you are going to see th difference between the 18-55 and the 21.

Lens that have excellent microcontrast and produce rich or striking color often seem sharper, because they produce images with more "snap" or "pop." So I suspect that some people confuse measurable resolution with contrast and color rendition. To judge a lens, one has to look at the images it produces, rather than how well it resolves MTF charts. (How many people buy a lens to shoo MTF charts?) It's the quality the images a lens produces, not how well it scores on numerical tests, that is important.

I agree guys...and while I think that DA 18-55 WR is pretty good lens, the limited lenses have this kind of magic (just more microcontrast, flare resistance and colour rendering) over it
and DA 21/3,2 looks really really good on K5 with the hood on

[luftfluss]

I'd say the 18-55 is *as* sharp as a lot of lenses *subjectively*, because I think sharpness is one of the most common features of any lens. I've used some real junk-bin lenses that were *sharp* (they wern't anything else, but they were sharp).

I honestly believe that sharpness is the most overrated aspect of lenses (likely a holdover from the yesteryear when sharpness actually was quite variable).

Sure there are a few lenses that are *VERY* sharp, but for the other 90% of lenses, the differences in sharpness do not matter at all.

Now when it comes to colour, DOF control, contrast, flare resistance... there is where you are going to see th difference between the 18-55 and the 21.

I agree that sometimes we obsess too much over the sharpness issue.

WRT the 18-55 lens, I've found it suitably sharp for taking pictures at fairly close range, in fact only very slightly inferior sharpness to my F 35-70 or M 28/3.5. However, for photographing more distant subjects (such as a landscape) my 18-55 falls down compared with the other 2 lenses. Foliage is not rendered nearly as detailed as the other 2 lenses. This I can notice with only a cursory glance at the photos.

[Biro]

We obsess too much about sharpness... and way too much about test charts and numbers. These can be useful, of course, but carry way too much weight with many people. Real-world results should carry the most weight of all. On that basis, my DA21 is a home run.

[civiletti]

I disagree to some extent with many of the comments in this thread. My take:

- The DA21 retest results on the k-5 are not stellar and not terrible. The DA18-55 has lower resolution at 18mm and higher resolution at 28mm, suggesting that it likely matches the DA21 at 21mm. The DA12-24 would likely test better than either.

- Sharpness [meaning resolution, not sharpness that can be altered in post-production] is not over-rated. It is one of the few optical parameters that cannot be enhanced much in PP. We can increase color saturation and tweak color balance. We can increase both overall and micro-contrast. We can correct linear distortion and remove chromatic aberration. In an image where maximum detail is desired, there is no substitute for a lens with high MTF50 results. If creamy bokeh is your nirvana, that's a different kind of image and only central MTF50 is critical, mostly at wide aperture.