Camera Labs

Here is one example of the effect, shot at f/2.8:


Nikon AF-S 35/1.4G 29131 and Nikon AF-S DX 35/1.8G 29123 by Thomas, on Flickr

The EV-delta in this case is around 0.7EV (in the middle, discounting vignetting).
Pretty crass, eh?

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

Thats really strange!

So I guess your next task is to bang on another lens and see if the 1.8 is cheating us or the 1.4 is just being extra generous?

_________________
Nikon D5000 and D300 with 12-24mm, 35mm f1.8, 50mm f1.4, 60mm f2.8 Macro, 18-200mm, 17-55mm f2.8 (all Nikon)

Already did that with the 35/2.0D, Jeremy!
And what can I say: That lens delivers even more light (around +1/4EV on average) compared to the 35/1.4G. Which sounds sort of OK, as the 35/2.0 is a pretty simple lens-design with only 6 lenses, while the 35/1.4G has 10 lenses inside.
So comparing a 35/2.0D vs a 35/1.8G would bring the gap to 3/4EV at least...

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

Btw. I think I can also dispel the suspicion that my D300 is botched: Just analyzed some shots from earlier this week on a D700. Same picture here: first tests showed a case of -0.7EV underexposure of the image from the 35/1.8G as compared to the 35/1.4G at f/4.0, 1/1000 sec or equally at f/2.8, 1/2000 sec.

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

It's certainly an unexpected issue to come across, but then lenses not meeting their quoted specification may happen more than you think.

It's just that most people rarely have a chance to compare multiple lenses with the same focal length at the same apertures under the same conditions moments apart. Most are tested separately in studios with the images adjusted in brightness to match for publication. Most would just also assume that if two lenses share the same aperture, then they'll deliver the same light.

The same applies to coverage. When I've compared various Nikkor kit lenses side-by-side at 18mm, they all deliver slightly different fields of view.

I've also seen it on actual bodies which claim to offer double the sensitivity as the ISO doubles, but actually the image brightness gradually decreases at higher values.

What's quoted is not necessarily what you get. Again we may rarely notice because most tests either don't look for it, or there simply aren't multiple versions of a lens with the same focal length to compare.

This is again while I support our testing policy of retesting multiple products side-by-side, moments apart, in order to make direct comparisons under the same conditions. You often notice things which are over-looked in studio-based tests which may be conducted on products weeks, months or even years apart.

I look forward to seeing how the aperture issues Thomas has come across plays out. It may be down to particular samples, or it may be a bigger issue that affects all models. I suspect the reality of optical design means the quoted values are often approximations, and technically may vary compared to the 'same' values on another product. Let's see...

To not break the Review announcement thread, where was said there is a low availability on the 35/1.8 DX lens, it could be the line is in use for a different lens at the moment, and if so, I would be glad to see the addition of more DX wide angle primes.

_________________
- Wout -
Nikon D80 with 18-70 DX, 70-300 VR, 35 1.8 DX
Manfrotto 055XPROB tripod with 488RC0 midi ballhead
Lowepro CompuRover AW and Nova 200 AW

Dear Thomas

Now, this is interesting.

Did the 35/1.8 have the same relative darkening at maximum aperture (from your tests, maybe it should). I wonder if it is the characteristic of the optical design of the lens or just the diaphragm. This would affect metering, because (I am assuming) Nikons meter at open aperture, do they not?

The other facor I was thinking of is how long the aperture blades take to close - slower closure would dim the image a bit, but would be less noticeable at longer shutter speeds. Although, if the lens is expected to be capable of 8fps, it should be pretty quick.

On the whole, I would have thought that the relative light loss at each lens surface would be quite small.

Finally, would there be some difference with the 35/1.8 because it is a retrofocus design but not computed as a wide-angle lens?

_________________
HCC
Nikon DSLRs, film cameras from Leica to Linhof

"Nikons meter at open aperture": Yes!
"the relative light loss at each lens surface would be quite small": Well, you may be right, but: there is more to transmission than the reflectance of the surface, which can be reduced with relative ease. There may be lens-elements that don't let as much light through as others. This is perhaps the reason for the 35/1.8 being so sharp and so cheap: some elements that have low transmission but otherwise ideal performance. But this is pure speculation.
"same relative darkening at maximum aperture": Yes indeed. That also indicates that the aperture blades have nothing to do with it. Only that I observed that at smaller apertures (f/8 and beyond) the irregularities seem to increase on all lenses.

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

Dear Thomas

Fascinating.

A couple more thoughts occurred to me - would the entry/exit lens size (ratio) have anything to do with it? I might see if I can establish a ratio on the weekend. I'll do some comparisons with my lenses, although that would only establish a relative value.

The point I made about the non-WA retrofocus design was prompted by the rather small front element - perhaps it has somewhat reduced light gathering power.

Normally, one would expece all f1.8 lenses to transmit roughly the same amount of light, but your point about the transmission qualities of different glass is interesting. I have always thought that the f-numbers have an almost direct relationship with the light but, really, they are a mesaure of the size of the opening of the lens relative to its focal length, and there are a number of complexities there (particularly with retrofocus designs, one would assume). Would this indicate an opportunity for the non-mirror cameras to have superior optice in certain cases (WA lenses come to mind)?

_________________
HCC
Nikon DSLRs, film cameras from Leica to Linhof

No. The transmission is to my knowledge only influenced by the number of glass-elements, the reflectivity of the glass-air-surfaces and the transmission of each element. So a lens with a low number of elements and a good coating normally has the better transmission-values.

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

Thomas

My apologies - I was referring to the possibility on non-retrofocus lens design for cameras without the need to have a swinging mirror (ie easier to design WA lenses). As a by-product, there could be less elements, but that's starting to get specialised for me.

Out of interest, that was one of the reasons for the Tessar type being more popular than the Gauss types in the first part of the C20th - no lens coatings in those days, and 4 elements/3groups did rather better than 6 elements/4 groups, all of which lends weight to your arguments. Even with coatings and specialised glass, some modern lenses make the light go through a lot of elements.

_________________
HCC
Nikon DSLRs, film cameras from Leica to Linhof

Well, there it is: The official Camera Labs comparison
Have fun!
---
Und diesmal auch in Deutsch: Der offizielle Camera Labs Vergleich.
Viel Spaß!

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

The best 35mm could be the Zeiss 2.0.

The zeiss 35mm 1.4 is also coming very soon.

_________________
Pascal Riben
http://www.pascalriben.com

The Nikon 35/1.4G is as sharp as or even sharper than the 35/2.0 Zeiss.
The 35/1.4 Zeiss is expected to fall behind.

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

I don't think so : zeiss lens have always been better than nikon lens.

_________________
Pascal Riben
http://www.pascalriben.com