15 replies to this topic

[mst]

If you're a regular visitor of our forum, you probably read the discussions following the reviews of the Nikkor AF-D 18-35mm lens on both the D7000 and D3x.

Quite often when we publish a review where the verdict doesn't quite meet the expectations of the community, we're accused or at least suspected of having measured a bad sample. Usually this is not the case, since we carefully check for several "alarm signs" during the review process, which help to identify lenses with issues. In fact, a good portion of the review work we do never leads to published results because of optical flaws we discover.

Well, I'm really sorry to report that in case of the Nikkor AF-D 18-35 indeed a bad sample was used for the reviews. For once, our triggers didn't set off the alarm bells and the poor measurement results led to verdicts that didn't do the lens justice.

Our reader and forum member studor13 kindly provided a second sample of the lens (thanks a lot!), which performed slightly better on FX, especially at the end, and significantly better on DX. I have updated both reviews accordingly, the optical verdict on FX has been increased slightly (from 1 to 1.5 stars) but quite a bit on DX (from 1.5 to 3 stars).

Since the initially tested lens was also used for the older D200-based review, which cannot be redone (due to lack of the test camera), I have removed that review.

The two remaining reviews can be found here:

FX: http://www.photozone...orafd18353545ff

DX: http://www.photozone...orafd18353545dx

-- Markus

[mst]

Forgot one thing: I have also shot some sample images with studor's lens. I'll update the gallery in the FX review tomorrow.

-- Markus

[you2]

Vaguely I thought you tried two samples originally and both performed badly (was that the case?); any conjecture as to why the other lenses performed so poorly? Mostly my motivation for asking this question is that I know some companies (olympus, zeiss, ...) will quietly change an optical formulae over time (no clue if Nikon does this sort of thing) and I thought this was an older lens in which you originally used a newer sample.

Forgot one thing: I have also shot some sample images with studor's lens. I'll update the gallery in the FX review tomorrow.

-- Markus

[BG_Home]

Thanks for the efforts, Markus. Though the results changed somewhat, your initial conclusions are still valid - poor on FX, and not really better than the competition on DX...

[mst]

Vaguely I thought you tried two samples originally and both performed badly (was that the case?)

I often base reviews on more than one sample, however that wasn't the case here.

any conjecture as to why the other lenses performed so poorly?

Not really. Maybe a combination of a bad element with some decentering. However, the first sample didn't show the usual signs of decentering.

I don't think that the optical design of the lens has been improved over time. Studor's copy of the lens performs more in line with the MTF charts published by Nikon (which, to my knowledge, remained unchanged over the life cycle of the lens).

-- Markus

[mst]

Though the results changed somewhat, your initial conclusions are still valid - poor on FX, and not really better than the competition on DX...

Yes, the general impression more or less remained, even though the results on DX improved considerably.

-- Markus

[PuxaVida]

Thanks for the (additional) effort Markus...

But I'm not sure if it's the good news: good sample gets 1,5 stars... If this was as good as it gets on FX, I would rather live with 1 star score and doubt for the rest of my life that it was a bad sample ...

Serkan

[TheChris]

Thanks Markus for updating the review. This kind of transparency and self-reflection is the essence of what makes photozone's tests so valuable.

Christian

[Studor13]

Firstly, thanks Markus for the report.

Before anyone gets too carried away with the 1 1/2 star on FX, remember that it's on the D3x and not on a D3. Hands up those who actually have a D3x and can't afford something more professional than the 18-35mm?

For the casual FX (D3/D700) shooter there is no other affordable wide-angle Nikkor zoom!

Also, note that the center performance at basically everything is more than decent.

BTW, how much exactly is the center?

When I started all this off I mentioned that David Ruether had many years ago said that even on film the corners were not great at 18mm.

However, if you read his report in full you can not be but fairly impressed with the lens.

http://www.donferrar...er/wa-zooms.htm

At 24, 28 and 35mm David gives either 14 or 15 out of 15 for center performances!

And this is his definition of center:

"The numbers given are ratings for sharpness for the major part of the frame, and in parentheses, the ratings for the far corner sharpness (the small area that would almost slip under a slide mount or negative carrier)".

Anyway, I think that it's all a bit academic as in the real world this lens is universally considered to be a good lens.

I think that it is always a good idea to get a number of opinions before making any conclusions about any lens.

For example, SLR Gear have an interesting way of testing and evaluating lenses since at the end of their report users can add their own comments. You will notice that 7 people contributed and not one disliked the lens.

http://www.slrgear.c...uct/113/cat/all

Lastly, in comparison to the 16-35mm, my initial findings are that the 16-35 is just better at everything except at 35mm (close call at 28mm). For some reason my sample has some issues at this focal length, so my workhorse for the past 5 years will not be quite retired just yet.

[Dreticus]

Markus, assuming that the current results are more representative for the actual lens population (which I don't doubt), it's good that you redid the tests. But I believe that you should mention somewhere in the reviews that this was your second copy, and that the results of the first were withdrawn because they were subpar. That's how Klaus does it in his tests. The addition of the word "revised" is a bit cryptic; and to me it gives the impression that your test method or your interpretation of the results was bad, while in fact it was not you.

[mst]

Yep, good idea. Done.

The gallery in the FX review has also been updated. Among better 35mm shots, it also includes a somewhat mean shot of a wall at 18mm wide open.

-- Markus

[Lomskij]

That brick wall shot is just... bad. Sums up this whole debate, I guess

On the other hand, if Canon indeed is going to release a new system camera with 2x crop, this lens might be ideal - center looks quite sharp

[Studor13]

"Sums up this whole debate, I guess". Really? I think you may have missed a couple of important aspects of this debate.

"That brick wall shot is just... bad." Yes, this lens has never been known to get good results when taking brick wall shots at 18mm and f3.5.

Try using this lens for other purposes that it was designed for. Let's say like the shot below (taken this morning).

[soLong]

ggrrrrrr

[PuxaVida]

Nice picture... But cropped (both the image and the sensor ) ...

But regardless of that... Why can't you be just happy with the lens and forget the 1,5 stars. If I had such a lens which I'm very satisfied, I'd never complain about the 1,5 stars. Testing a lens can never 100% mimic the real life usage (and I guess this is known by the testers already). Markus and Klaus just try to quantify the results of a system so that a holistic approach and the results for that particular lens could make sense at the same time. This is not an easy thing actually...

Serkan

[Studor13]

"Why can't you be just happy with the lens and forget the 1,5 stars"

I am happy, believe me.


"... But cropped (both the image and the sensor)"

Just for you. 18-35mm at 18mm on Velvia.


.


.


.