16-35 vs. 18-35

[simsurace]

Another vote for the 18-35.
Very happy with mine 

Thanks!

[richardHaw]

I see. Read some reviews and saw some images and they do look good as far as i can see

[richardHaw]

They are also cheap, saw one here for ¥50,000 in like new condition with everything. I was however turned off by the lack of aperture ring...(macro purposes)

[simsurace]

I ended up buying a 18-35 two months ago. I already have made a couple of hundreds of shots, but I've never taken out both the 18-35 and 16-35 at the same time. My impression is that the 18-35 does almost the same job, but it's much lighter and smaller.
Ok, today I went for a walk and made some comparison shots between the 16-35/4 and the 18-35/3.5-4.5.
All shots are on the D600 at base ISO and on a tripod. VR was off for the 16-35. I developed in ACR, CA correction was on, otherwise no lens corrections. Moderate sharpening (amount 40, radius 0. applied equally on all the shots.


Scene 1: Close focus, bokeh, backlight
Because the physical length of the lenses is different, I ended up shooting a different focal length, but the shots are fairly similar and have the same aperture.
18-35, 35mm, f/5.6, 1/350s


16-35, 28mm, f/5.6, 1/350s


Commentary: Very similar rendering, bokeh reasonably smooth, good contrast, no flare, decent sharpness.


Scene 2: Mid- to far-field, extended DOF, fine detail, corner sharpness

18-35, 20mm, f/11, 1/10s


16-35, 20mm, f/11, 1/10s


Commentary: Again, very similar rendering with minor geometric differences. Both lenses render nice colors.
100% crops from the upper right corners reveal slight differences:

18-35:


16-35:


Commentary: The 18-35 renders the corners of the image with more sharpness and less geometric distortion.


Scene 3: Infinity (or close), wide aperture corner performance.

18-35, 24mm, f/4, 1/500s


16-35, 24mm, f/4, 1/500s


Crops from upper right corner:

18-35


16-35


Commentary: Again, the 18-35 delivers more sharpness in the corners of the frame.


Scene 4: Distortion

18-35, 30mm, f/11, 1/8s


16-35, 28mm, f/11, 1/8s


Commentary: Neither of the lenses is distortion-free around 30mm. The 18-35 shows barrel-type and the 16-35 shows pincushion-type distortion. In general, both lenses have distortions which are fairly easy to correct. The 18-35 is always barrel, the 16-35 changes from barrel to pincushion and is almost distortion-free somewhere close to 24mm, which is quite nice, but not tremendously important.


Interim conclusion: I'm not yet completely sure, but I'm leaning towards selling the 16-35 and keeping the 18-35. So far, I haven't encountered a situation in which the 18-35 disappointed me and for me the 16-35 doesn't really show many advantages in my type of shooting. The 18-35 impresses me with the sharpness even close to wide-open and in the corners. The contrast, color etc. is almost indistinguishable from the 16-35. Moreover, it is smaller and lighter, which makes it easier to carry and use.

Any other observations? Thoughts?

[Frank Fremerey]

I have a clear understanding that the added features like VR or shorter focal length in wide go with decreased IQ.

So:

Either get a good copy of the 18-35 or the real thing 14-24 or a GREAT compromise 17-35.

Difficult choise I know.

[bjornthun]

The corner performance of the 18-35 @24 seems definitely better than that from the 16-35. My guess is that the 18-35 AF-S will be the better landscape lens of the two. The AF-S 18-35/3.5-4.5 actually looks to be a very nice lens.

I'm actually a bit surprised that a DSLR lens shows corner smearing like this, since it should have no problems with an oblique ray angle at all. There are gode WA zooms with image stabilzers built in (Canon & Sony) so that's no excuse for the Nikkor 16-35/4.

[Bjørn Rørslett]

Must be sample variation? I had a 16-35 for review and it was sharp all over the frame. No corner smearing. Also had decent IR and VR performance.

The latest 18-35 incarnation also was a very good performer. I had difficulties telling the two apart.

[simsurace]

I have a clear understanding that the added features like VR or shorter focal length in wide go with decreased IQ.

So:

Either get a good copy of the 18-35 or the real thing 14-24 or a GREAT compromise 17-35.

Difficult choise I know.

I don't know, the VR and shorter focal length also come with additional bulk and presumably better glass. I would not have expected the result.

The 14-24 does not fulfill my requirements for a light and small solution.

The corner performance of the 18-35 @24 seems definitely better than that from the 16-35. My guess is that the 18-35 AF-S will be the better landscape lens of the two. The AF-S 18-35/3.5-4.5 actually looks to be a very nice lens.

I'm actually a bit surprised that a DSLR lens shows corner smearing like this, since it should have no problems with an oblique ray angle at all. There are gode WA zooms with image stabilzers built in (Canon & Sony) so that's no excuse for the Nikkor 16-35/4.

Yes, it does.

My understanding is that corner smearing would be an issue if the sharply projected image suffers due to an angle of incidence which is too shallow. Here, it seems that the image is already smeared, either because of decentering or by design.

Must be sample variation? I had a 16-35 for review and it was sharp all over the frame. No corner smearing. Also had decent IR and VR performance.

The latest 18-35 incarnation also was a very good performer. I had difficulties telling the two apart.

My first thought after getting the 16-35 and seeing the corner performance was that it must be due to some defect. But Nikon told me that everything is within spec. I thought I was expecting too much from a wide angle zoom. But now the 18-35 shows that more is possible.

I also have difficulties telling them apart from the entire image, but with my samples, close inspection gives away which is which.

[bjornthun]

"Within spec" must have a very liberal definition nowadays. 

[MFloyd]

I own a 16-35mm f/4 VR which is "perfect" for my work environment i.e. photography on "moving" platforms such as boats and planes.  The choice for me was between the aforementioned and the 14-24mm; the latter front lens being too exposed for what I'm doing. With regard to the quality: almost perfect to my standards; VR is for me a must.

I'm definitely not a believer in varying quality, given the quality control process; but this is another debate.

[Frank Fremerey]

The 18-35 has a very serious weight and size advantage over the
other candidates and is is also quite cheap.

What MFloyd says means for me that faults detected in a
sample you buy can and must be corrected by Nikon for new
items or within a certain amount of time free of charge.

My new 1.8/50G was seriously decentered. They fixed it.

For older and used items it is more tricky because the item
price is often smaller than the fixing costs. Also Nikon can not
be held responsible for a lens biography with possible mishaps.

[Bjørn Rørslett]

...
I'm definitely not a believer in varying quality, given the quality control process; but this is another debate.

My best estimate of failure rates for Nikkors is < 2%. This figure is based on testing hundreds of lenses over the years. Typical failures are decentering from misaligned lens elements, excessive dirt inside the lens assembly (more cosmetic than a true annoyance, though), or communication errors between camera and lens.

I have had a lens (15/5.6)  so badly decentered that the factory could not get the lens into acceptable shape even after several trips back to Nikon Japan. An AFS 400/2.8 straight out off a sealed factory box couldn't focus at all and probably had a flipped element somewhere inside. Zoom lenses with decentering issues are not unusual, leading to very different performance in either end of their focal range, or part of the frame being unsharp. The last time I observed this was with a brand new 70-200/4.

I'm not saying Nikkors are badly made, not at all, but denying the existence of quality variation is futile. It certainly exists.

Quality control these days is different from way back in time. Now the production relies more on proactive methods, viz. making designs that are much less sensitive to the inevitable component variation, and letting the customer be the final judge for quality.

[MFloyd]

@Bjørn: off course, I have not your in-depth sight. 2% being small, but big on the other side; probably the best ratio in economic terms. 

[Bjørn Rørslett]

I used the figure '<2%'. Probably lower on the expensive low-volume stuff and might be higher on the mass volume items. The estimate is as good as I can get from own observations.

The important aspect is that the failure rate is *not *zero.

[simsurace]

I own a 16-35mm f/4 VR which is "perfect" for my work environment i.e. photography on "moving" platforms such as boats and planes.  The choice for me was between the aforementioned and the 14-24mm; the latter front lens being too exposed for what I'm doing. With regard to the quality: almost perfect to my standards; VR is for me a must.

I'm happy that your 16-35 is serving you well!