[rawr]

DxOMark have tested and reviewed the DA 20-40 WR:

Pentax HD DA 20-40mm f2.8-4 ED Limited lens review: Prime performance in a zoom - DxOMark

They seem to have liked it - tests look OK. Maybe they got a better sample than Lenstip

[VisualDarkness]

Seems pretty comparable to the other brands offerings. Though a bit more expensive but probably just like with Zeiss you pay a lot extra for just the metal build.

[Mr Spocko]

I trust lenstip most of the time their findings are usually in line with my real world shooting. DxO well numbers mean little to me, I like real world images to see what's going on.
It is possible to get sample variation no doubts there. It will be interesting to see more reviews of this lens, so far I'm not convinced with what I've seen

[JinDesu]

For sharpness reference using what DXO shows, the DA20-40 registers 9MP on the k-3, while the DA 35 F2.4 registers 11MP, the DA12-24 is 10MP, and the DA 18-270 is 7MP.

[lytrytyr]

Maybe they got a better sample than Lenstip

Actually, the quantitative results from Optyczne/Lenstip are very good for a zoom,
e.g., an edge resolution stopped down that's well above 90% of the FA 31's.

But like many commentators on this lens,
Arek doesn't seem to comprehend the specific size/performance/cost compromise
that underlies the Limited lens philosophy.

[VisualDarkness]

For sharpness reference using what DXO shows, the DA20-40 registers 9MP on the k-3, while the DA 35 F2.4 registers 11MP, the DA12-24 is 10MP, and the DA 18-270 is 7MP.

I have a tough time with DXO most of the time but that list sounds very realistic.
The DA 35/2.4 is supposed to be really good optically, but kept cheap due to a rather conservative design and plastic build.
DA 12-24 is considered to be one of the sharper extremely wide lenses but with some CA troubles (as most Tokina designs).
DA 18-270 performs like a super-zoom, nothing more and nothing less.

I can see and accept the 20-40 right in the middle there when it comes to sharpness.

[lytrytyr]

For sharpness reference using what DXO shows, the DA20-40 registers 9MP on the k-3, while the DA 35 F2.4 registers 11MP, the DA12-24 is 10MP, and the DA 18-270 is 7MP.

In terms of the DXO metric,
"Best resolutions are usually attained for fixed focal lenses and moderate apertures."

Their metric is a meaningless figure having little to do with actual lens use,
especially for the Limited zoom
which sacrifices low-aperture wide-angle edge sharpness for small size
(comparing to the Sigma 18-35/1.8, for example).

You could easily game this meaningless metric by limiting the aperture settings to f/4-f/11.

---------- Post added 02-19-14 at 05:53 PM ----------

Though a bit more expensive but probably just like with Zeiss you pay a lot extra for just the metal build.

As Mitakon proves, metal build doesn't make a lens expensive.

[VisualDarkness]

As Mitakon proves, metal build doesn't make a lens expensive.

Nor do Voigtländer but that doesn't hinder manufacturers to take a premium for solid metal builds.

It's a bit of "How few lenses can we sell but still make maximum profit as a whole?" kind of deal with the premium lenses.

[JinDesu]

In terms of the DXO metric,
"Best resolutions are usually attained for fixed focal lenses and moderate apertures."

Their metric is a meaningless figure having little to do with actual lens use,
especially for the Limited zoom
which sacrifices low-aperture wide-angle edge sharpness for small size
(comparing to the Sigma 18-35/1.8, for example).

You could easily game this meaningless metric by limiting the aperture settings to f/4-f/11.

---------- Post added 02-19-14 at 05:53 PM ----------



As Mitakon proves, metal build doesn't make a lens expensive.

I don't typically like DXO's system, but I wanted to give a bit of comparison information that helped to understand where the DA20-40 stood amongst other common Pentax lenses. And I also looked at the field maps for sharpness to be sure that the values assigned to the 4 lenses mentioned made sense. In the field maps, the DA20-40 did very well in the entire zoom range from F2.8 up to about F11. So it's not a bad lens, based on those metrics.

That being said, it's only a measure of sharpness and yes - we can easily game the metric by putting the Sigma 18-35 at F4 and seeing it crush the DA20-40 at F4, but I am assuming that people on this forum are intelligent enough to understand that the sharpness comparison between familiar Pentax lenses is only to help put one parameter into place. With regards to actual lens use, they can refer to other sources including Lenstip, which describes a whole medley of parameters like flare resistance, coma, vignetting, and build quality etc.

[lytrytyr]

That being said, it's only a measure of sharpness and yes - we can easily game the metric by putting the Sigma 18-35 at F4 and seeing it crush the DA20-40 at F4

Actually, what I meant by "game the metric"
was that the DA 20-40 score would be boosted
if you restricted consideration to the f/4-f/11 range.

For edge sharpness, the Sigma zoom is comparable to the FA 31,
so only a few percent better than the DA 20-40 in its comfort zone at f/8.

The flare resistance of the DA 20-40 is very much superior,
making the lens a joy to use in strongly back- or side-lit situations.

[Pål Jensen]

The flare resistance of the DA 20-40 is very much superior,
making the lens a joy to use in strongly back- or side-lit situations.

Thats why I will buy it cause the 16-50/2.8 simply isn't good enough in this regard.