[mst]

Unsurprisingly, very nice on DX, too.

 

http://www.opticallimits.com/nikon--nikk...7020028vr2

 

-- Markus

[stevenb]

A nice big and heavy lens. To big and to heavy and to long for me .

[JJ_SO]

In another test, Markus, you asked if it's worth to buy 85/1,4G while 85/1,8G gives you 95% of it's performance at 1/3 of the price.

Same question for 70-200/2,8 VR II vs. 70-200/4 VR (III): I tried a friend's f/2,8 and found thenVR of the f/4 is able to equalise the lack of speed to a certain point. Of course, if one needs faster aperture because of sport / bird / wedding shooting, it has to be the expensive one. Resolutionwise, they are very close.


And although I could have afford the more expensive f/2,8, I'm a happy user if the more modern, lighter and smaller f/4.


So, in my eyes the test is a bit outdated: both FX and DX users have these days sensors available with 50% more resolution and faster AF, at least the D7000 vs D7100 is more sensitive and a bit faster, thus not more precisely. I see the reason to get comparable results, but this comes with open questions how those lenses perform with new cameras?

[Klaus]

We are not talking about a 50% difference but 22% in term of LW/PH.

FWIW, in psychology a difference of 20% in a row is considered to be just recognizable.

 

However, as mentioned Markus has already the D7100 in his stock. This is basically 'old data'.

[JJ_SO]

I am talking about 50% more information theoretically and don't care about other math games. But I see your point and was not able to recognise a 50% quality jump on normal pictures. Talking psychologically, those MP are looking worse than D7000 if looked at it in 100% enlargement. But on full screen at high ISO it's a whole different story. And not depending on the lens in front of the camera.


But alright, if it was "only" cooking old data...


But even then my question remains how to see and judge the difference between both zooms?

[Klaus]

This doesn't change the fact that the D7100's LW/PH figures will just peak about 20% higher than for the D7000 - and the qualitative difference will be marginal.

[mst]

I don't think one can compare this case to the debate over which 85mm prime to get. If one needs to decide betwenn f/1.8 and f/1.4, speed is usually not the deciding point. At least in case of the 85mm primes, the main advantage of the faster lens is better bokeh and better build. But the differences are marginal, IMO.

Talking about tele zooms, the step up from f/4 t f/2.8 is significant, and there are several reasons to go for the faster lens, as you already mentioned. There's better AF performance, narrower DOF, a brighter viewfinder image, TC compatibility... while on the other hand, the faster lens is no doubt more expensive and a lot bigger and heavier.

Usually, the camera used is a given. So the question is, which lens to get for my camera, not which lens/camera combo to buy.

Apart from that, as Klaus alread mentioned: I already started testing with the D7100. I will nonethelss continue to publish a few more D7000 based review for a while, so that the results are comparable.

-- Markus

[Guest]

I have this lens, I bought it 2.5 years ago.  I can confirm it's excellent performance.

 

If I were buying now, with the knowledge I have, I would buy the f/4 version.  MST's tests show a comparable performance, and obviously price is a factor.

 

The f/2.8 is not just 'heavy', weight is concentrated towards the front of the lens by those big elements and instant backache is the result. Even when balanced by a D200 with battery pack and two batteries.

 

The result of this is that the lens usually stays at home. I know that the f/4 would travel, as I have the 24-70 f/2.8 Nikkor, I use it a lot, and the handling is about the same.

 

I guess I'm my own worst enemy because I accumulate stuff and never dispose of anything, so it's a keeper - but it does not get the use it should.

 

There is an interesting article from Lensrentals in the USA, you can get to it through the current DP Review home page.  A This is most likely spam content of hundreds of rented lenses shows failure rates. The various 70-200 lenses - current models - fare badly, as a breed.  No particular reason is suggested, but the sheer complexity of them is pointed-out. No manufacturer covers itself with glory here!

 

Best wishes to Zoners. Even the upside-down ones.....

[JJ_SO]

In only can agree, AAC7man, to the things you said to the f/2.8 version and the f/4 version. Just came back from a 2 week cycling trip on the rough roads of Balticum - Lithuania and Russia. The f/2.8 version would clearly had a stay at home, while the f/4 was good to have with me and never a problem of lacking speed - only forgetting to switch on VR which was OFF sometimes because it needs power and when travelling abroad I try to switch off as much energy consuming features as possible.

 

I understand the weight difference: bigger glass elements, a lot of solid mechanics, metal tubes and not much lightweight plastics on it. So, if there's a problem lateron, repair costs will also be high, just because the mechanics are complex and need every bit of precision.

 

Here's the link to the repair report: Lensrentals Repair Data: 2012-2013

 

But as Roger said, one should keep in mind, those lenses are rental stuff and in much worse (ab)use than your own would be, so one has to interprete carefully the mass of data they collected.

 

I learnt, it pays sometimes to wait with buying fast Nikon lenses. While the 85/1.4G is great, it also has the flaw of being a heavy lens and repairwise, it might be a disadvantage the HSM has to move a big mass of glass, compared to it's less faster brother f/1.8. Today I would go for the less expensive one.

 

But now: Selling this lens and getting a less faster is something I hesitate to do. That would become a different story, if Sigma relleased one day a 85/1.4 A - because one of the downsides is CA and bokeh fringing of that expensive lens and it could be that this Sigma will become superior. But I even didn't hear any rumours so it might take a little while.