[leopold]

Civiletti,
if you're interested i posted a short review also with some crops .... sure not test charts but real life photos :

PENTAX DA 20-40mm F2.8-4 ED Limited DC WR review. | steevemarcoux2

[lytrytyr]

Possibly because Adam did careful testing.

There were some issues with the test, see
https://www.pentaxforums.com/forums/10-pentax-slr-lens-discussion/247626-da-2...ml#post2639479

However, the main problem seems to be
that Adam's copies came early in the production process.

[lytrytyr]

A few exposures posted at common webpage size prove little about lens performance.

Crops work fine on the web, e.g.
https://www.pentaxforums.com/forums/122-lens-clubs/247859-da-limited-zoom-club.html#post2642014

[Zygonyx]

Hopefully we should soon find out others reviews, with real measurements.

[cali92rs]

Hopefully we should soon find out others reviews, with real measurements.

That won't matter...if it measures bad, people will say "it just renders beautifully".

[Kenntak]

Why can't people simply enjoy what their lenses produce, without having them subject to scientific testing for justification?

[thibs]

That won't matter...if it measures bad, people will say "it just renders beautifully".

If it measures well ,you'll be first on the line to claim that this lens makes no sense anyway.
Finished?

[cali92rs]

If it measures well ,you'll be first on the line to claim that this lens makes no sense anyway.
Finished?

I wasn't first in line...but it sure was a long line

[Dice]

I've come to understand that people will see what they want to see. Those that wanted something faster, something smaller, something wider, something longer, will not be happy no matter how good the lens is and will focus on anything negative. As with any human manufactured, designed device there will be sample variation. Whether that was an issue with Adam's review I don't know. For those who are interested and want to consider another viewpoint that this zoom should be considered if you like the focal length range. This tester generated imatest data for the DA21, DA35 ltd, DA40, and the DA20-40 all on the same K-3 camera. If compare the data from each review you have another piece of information to base a decision as to whether the lens should be able to produce nice sharp images across the focal range. As having come from all of the compared DA primes, I can offer my unscientific assertion that yes it equals or betters the DA primes.

That won't matter...if it measures bad, people will say "it just renders beautifully".

Measurements don't matter to the degree that I don't normally take pictures of line tests. At the end of the day, I buy and hold onto lenses that do what I need them to do. Not because they scored high on any specific test. It just so happens I do value sharpness, WR, versatility all in a compact package so the 20-40 will stay in my bag and the DA21, and DA40 are gone....and I'm wrestling with selling the DA35 ltd. And by the way it does render an image beautifully...at least to my eye.
YMMV.

[Parallax]

I wonder why people judge how good a lens is by how the picture looks at screen resolution. That's roughly the equivalent of judging a lens in the pre-digital days by how a huge poster size print looked. Viewing the entire image (fit to screen) on 21" monitor from an APS-c sensor (Slightly more than 1" Diagonal) is a 21x enlargement, or roughly the equivalent of viewing a print of 12x18. A print from a 35mm negative (roughly 1.5" diagonal) at the same 21x enlargement would be about 18x27. The "standard" was typically 8x10 (or 8x12 for 3:2 purists), an enlargement of only 8x, so even without pixel peeping we're judging lenses at more than 2.5 times the magnification that we used to. Pixel peeping is like looking at a print from a 35mm negative of somewhere around 60x40 inches, or more.
Was a lens that could produce a tack sharp 8x12, but was a wee bit fuzzy in the corners on a 60x40, a sub-par lens?

[Zygonyx]

Thanks for links to reviews

[Cannikin]

I wonder why people judge how good a lens is by how the picture looks at screen resolution. That's roughly the equivalent of judging a lens in the pre-digital days by how a huge poster size print looked. Viewing the entire image (fit to screen) on 21" monitor from an APS-c sensor (Slightly more than 1" Diagonal) is a 21x enlargement, or roughly the equivalent of viewing a print of 12x18. A print from a 35mm negative (roughly 1.5" diagonal) at the same 21x enlargement would be about 18x27. The "standard" was typically 8x10 (or 8x12 for 3:2 purists), an enlargement of only 8x, so even without pixel peeping we're judging lenses at more than 2.5 times the magnification that we used to. Pixel peeping is like looking at a print from a 35mm negative of somewhere around 60x40 inches, or more.
Was a lens that could produce a tack sharp 8x12, but was a wee bit fuzzy in the corners on a 60x40, a sub-par lens?

Sorry, but your analogy is off there, as prints have much higher resolution than the vast majority of screens out there, and can (and usually will) be viewed much closer than people feel comfortable looking at a screen of comparable size.

Good prints are made such that viewed closely, a person without visual aid cannot easily distinguish the dots that make up the image (standard quality being ~300dpi). A standard 1080p screen has a whopping 2 megapixels (actually a little less if you want to display a 3:2 image on a 16:9 screen without cropping). A 21" 1080p screen is hard limited to an abysmal ~100ppi resolution, worse if the screen is bigger as many people display on TVs nowadays. If you look as closely as many people view prints, you are seeing nothing but the limitations of the screen and learn absolutely nothing about the limitations of a lens.

Unless you have an 8K screen (~33MP), you will not be able to match 300ppi resolution on that little 21" screen, and cannot judge an image as closely as a print typically will be.

[Parallax]

If what you say is true, then why do photos taken with the 31Ltd look better on a monitor than those from the 18-55 kit lens? The 18-55 already out resolves monitors, so shouldn't the pictures look the same?

[Cannikin]

If what you say is true, then why do photos taken with the 31Ltd look better on a monitor than those from the 18-55 kit lens? The 18-55 already out resolves monitors, so shouldn't the pictures look the same?

Do they now? I just went out on my balcony and shot with the 18-55 WR kit lens and the FA 31 Limited. Can you tell which one is which? Both shot on the K-3 at f/8, focused at infinity (so no DOF/bokeh clues), and developed to JPEG in camera. Resized in Photoshop.

Lens A:


Lens B:


It's pretty blatantly obvious which one is the better lens at 100%, but can you tell at this resolution? You can download them bigger if you want, but I limited them to 1080 lines of vertical resolution, so no "cheating" by blowing them up bigger. Don't want to "unfairly" judge these lenses at huge magnification right?

[Parallax]

You just validated my point.