Nikon 70-200 f4

[Lance B]

I don't think it's a matter of my agreeing with your collective thoughts on TDP .. but what's pretty obvious when comparing TDP's actual images with other sites findings is that they're pretty much on par.

I can't understand how when a review site A seems to get similar results to review site B, review site A is not a trustworthy operator!
I'm afraid it's simply beyond my ability to see that as logical.

And it must also be remembered that not every site will get exactly the same results from the lenses they test compared with the test data from other sites, due to sample variation from each lens.
And while it's a populist approach to denigrate the likes of Sigma and Tamron

Populist to denegrate Sigma and Tamron? Really? The latest Sigma 35mm f1.4 has been shown to be a superb lens and outresolves the equivalent Nikon by quite a margin (a lens I have I might add) and the bokeh from the Sigma is also considered excellent and there is a flood of Nikon users showing outstanding results from it, and from what I have seen I agree, so you can't use that hoary old chestnut that "only propriety lenses get the accolades" excuse as that just isn't the case. When a top quality third party lens comes along, then it gets the deserved accolades.

among others for their higher levels of sample variance with their products, the issue is not strictly limited to these non primary manufacturer's!
Photozone, who I always keep a watchful eye on myself had an interesting review of the Nikon 70-200/2.8 VR(original version) .. where they tested multiple lenses of which a couple were good, or great and others which were .. quote .. "suffered from a rather pronounced centering defect".
On the original test for the 70-200VR, Markus only referred to testing two lenses, but in a subsequent post on another forum he went on to explain about 4 lenses being tested in total!

Can you link me to that info?

And this is a site that you do trust!

Well of course!! Because they took the time to make sure they obtained a good copy and ensure that the lens they tested reflected what could actually be expected of the lens. I don't find that strange at all.

I'm not entirely sure that a site should be trusted if take such liberties with one brand, but not extend the same testing procedure with others brands too!

How sure can we be that PZ's copy of the Sigma 70-200/2.8 didn't suffer from some unknown defect? Did they attempt to test another copy to be sure that it wasn't transport damaged, or something like that?

Sorry Arthur, but talk about paranoia. It's quite simple, the Sigma they tested didn't have the centring defects. They apply the same controls for the Sigma as they do all lenses and they have had to get better examples of other Sigma/Tamron/Canon lenses in othe lens tests due to decentring and other manufacturing irregularities. No need to think there is a conspiracy against Sigma or Tamron because it ain't.

Anyhow, as I already said, if you look at the data on TDP, it basically mirrors what PZ has found anyhow, with slight differences which can be explained via sample variation ... where the central resolution of the Sigma is approximately 93% of what the Nikon VRII lens is capable of .. and this is regarded as "doesn't come close in any department".
I'm not sure which dictionary this explanation comes from, but it's certainly not one that I ascribe too.(that one would be definitely deposited into a bin in my household)

In my book that isn't close when you look at other lenses tested and can easily be seen when viewing the difference in IQ. I have had both lenses of tests that have been tested on Photozone that show differences of less than 7% and that can be clearly seen. The Sigma 35mm f1.4 mentioned above will probably only have a resolution figure in the order of 5-10% more than the Nikon 35mm f1.4, similar to the difference you pointed out for the 70-200's, but to me that is a significant enough difference to plainly see. Again, it doesn't make the Nikon poor as it isn't, it is a superb lens, it's just not as good as the Sigma version and just like your Sigma 70-200 f2.8 isn't a bad lens, it's just not as good as the Nikon.

But of course the corners of the Sigma on PZ fail badly, as per their test.

Isn't this also part of my "isn't even close" comment?

But then they never spoke about decentering and testing another copy of that lens anywhere on that review page!
I have my theories on why this is.
(refer to my comment above about trust!)

Again, paranoia. Photozone has had to get decent copies of other brands lenses due to decentring and other irregularities. It's not a conspiracy.

Like I said, I trust only what I see and read, and make my conclusions based on what I believe is happening and subsequently referenced to other sites both known and unknown ... not what 'my mates and their mates' think this guy is doing and what his motives may be. I realise that there's an absolute glut of respected folks out there that known a few things about certain aspects of photography. And I'm not going to be one to argue with them about their ingrained ways .. they more than entitled to their opinions on their preferred sites!
My point is that I'm not closed to any site as a source of info, unless the info is obviously idiotic!!!
And one particular name springs to mind the moment i think of such as topic .... Kay .... Ahhhhh!
But even then despite all the lunacy of contradictory information on that site, it does have a few rare moment of normality and competence too(yep! hard to believe, but true)

Maybe so, but how do you know when to trust that particular site?

I can't see any reason to distrust The Digital Picture's testing based on the images they present as data for each lens.
Apart from an anomaly or two here and there with the actual images, which is most likely due to variance between lenses, I see very similar IQ and patterns of IQ across the frame with the mutual lenses I have in my possession!
Of course my test target is much smaller than theirs are, and I use mine simply as a first step to determine what action I need to prepare for with any new lens .. but once that's done the chart I have never gets used again for months or years.

Anyhow, the (slightly off)topic has been bashed to a pulp from my end, and now there's a new kid on the block .. but my mind is almost now set(about which 70-200 I'll end up with in a few months), even tho I haven't yet sen enough proof to warrant my decision.

If you are shooting Nikon, then I think you will find that the 70-200 f4 will be hard to beat if you take price into consideration and the fact that VR seems to be up to 5 stops of handholdability. Quite remarkable. Added to that, it works brilliantly with TC's, better in this regard than the highly regarded 70-200 f2.8 VRII. If I didn't have the f2.8 version, I would be very swayed by the new 70-200 f4.

[arthurking83]

......


Can you link me to that info?


......

I won't link the other forum, but it is a well known Nikon centric site.
But the info about the problematic 70-200VR is right there in the original incarnation of the review .. i.e D200 test.

[swifty]

Just adding to the confusion :P
Another option has just appeared in the form of a soon to be launched Tokina 70-200 f4 with stabilization. Tokina's also well reputed as a quality alternative to original manufacturer's lenses.
My personal opinion is you're splitting hairs in terms of IQ at equivalent apertures. If you need/want the f2.8 your penalty is weight and cost and your shooting habits will dictate which one's more suited. I personally have the f2.8 VRII version and its my most used lens. If I had the money I'd love the f4 in addition but it can't replace it for my line of shooting.
I've also had less success in the past with third party lenses but you can't draw conclusions from my one sample point. And despite this I'm still considering a Sigma lens (35mm f1.4 to be exact) as my next purchase so gone are the days where original manufacturer lenses automatically meant better performance so definitely consider third party alternatives as has been suggested, particularly if you're pushing your budget.
Good luck with your choice

[arthurking83]

Photozone has tested the 70-200/4.

As expected it's almost as good as the 70-200/2.8VRII ... not better, as some experts seem to think.

It seems that (according to PZ's comments) the VR feature is better than earlier incarnations of VR.
But I'd reckon it has more to do with the lower weight of the lens too.

[I @ M]

Looking at the photozone test it appears that the F/4 would be a perfectly logical choice for someone who primarily photographs hand held, the price point and IQ look very appealing. But --- as swifty notes, the Tokina might give it a big run for the money when it is released.

[Lance B]

I think the issue with all of these testing sites is that they only test at one subject to camera distance. Some lenses work better at longer camera to subject distances and some work better at closer camera to subject distances. It would be great if they could do 2, or even better 3, camera to subject distance tests to show where each lens' strengths and weaknesses are.

[arthurking83]

They only test different focused distances at various focal lengths and sensor formats used ... that is, the 70mm will be closer to the target than the 200mm test will be.
I think the idea is that they try to fill the focus chart to the framing.

This then means that a 200mm lens tested on a full frame camera is going to be tested at a shorter working distance if that same lens is mounted on a smaller senor camera(APS-C or 4/3rds .. etc).

Anyhow, the point is that these tests aren't supposed to be absolute. That is each test is not the definitive last word on the performance of the lens in every possible permutation and combination of each variable.
Anyone that supposes that their test is the last word about the performance of the lens is seriously deluded.
These tests are supposed to be an indication of what you could expect to see if you compare one lens to another. Expecting anything else is not a realistic proposition in the real world.

So, to perform the set of guidelines as you propose is both futile in the real world, and would end up a logistical nightmare to boot.
It's a futile effort in that all that testing will be made for no real gain .. my copy of that lens is almost certainly going to perform differently than your lens and their lens. So the point of their extended effort is what?
For sure more test data will be available, but is it really relevant, and necessary?
Anyhow ... it's simply not necessary .. and this is where opinion based reviews come in handy(such as this Brad Hill fella), and their more extensive findings.

There are many of these Brad Hill's in the world, and in Nikon circles(that I know) the most respected is Bjorn Rorslett, in both his analyses and methods(he gives simple scores) all based on testing in usage not focus targets. He explains if a lens is better at infinity or closer up, and so on and so forth. But his knowledge of technical aspects is pretty much unsurpassed.

[Lance B]

They only test different focused distances at various focal lengths and sensor formats used ... that is, the 70mm will be closer to the target than the 200mm test will be.
I think the idea is that they try to fill the focus chart to the framing.

This then means that a 200mm lens tested on a full frame camera is going to be tested at a shorter working distance if that same lens is mounted on a smaller senor camera(APS-C or 4/3rds .. etc).

Anyhow, the point is that these tests aren't supposed to be absolute.

I am fully aware of that and I never intimated they were, just that if they did use different focus distances that it would give a better indication of lens performance for people to get an idea of how it performs for a specific task that the potential purchaser may use it for. Take a macro. What's the use of doing a test at 50x focal length when it will be used at probably less than 5-10x focal length for most of it's life.

That is each test is not the definitive last word on the performance of the lens in every possible permutation and combination of each variable.
Anyone that supposes that their test is the last word about the performance of the lens is seriously deluded.

Never suggested that I ever thought they were the last word in testing as I have a great deal of experience with a vast array of lenses. However, "seriously deluded"? Awfully strong words regardless. What are you implying? I am deluded? Thank you for the heads up on that.

These tests are supposed to be an indication of what you could expect to see if you compare one lens to another. Expecting anything else is not a realistic proposition in the real world.

So, to perform the set of guidelines as you propose is both futile in the real world, and would end up a logistical nightmare to boot.

I don't see why and I completely disagree. It would not be difficult to not only test for the ususal 50x focal length and say 500x or 1000x focal length and then display two resolution figures for a prime and say 3 for a zoom. They do at least 3 tests for a zoom as it is, on the 70-200 it is 70mm, 135mm and 200mm, so what's wrong with doing a prime at at least 2 focus distances?

It's a futile effort in that all that testing will be made for no real gain .. my copy of that lens is almost certainly going to perform differently than your lens and their lens. So the point of their extended effort is what?

Well, I don't think that is correct. In this instance, it seems that there is enough anecdotal evidence that the 70-200 f4 does perform better at distance over the 70-200 f2.8 and visa versa. From what I have read about other lenses that I have and compared them to what others say about their versions of those same lenses is that mine behave very similar if not exactly the same way as theirs. So, I have to disagree when you say that each lens of the same type is that much different. If there was that much difference between lenses then these tests would be completely futile.

For sure more test data will be available, but is it really relevant, and necessary?

To someone who wants to know how their lens behaves at distance rather than a test chart at 50x focal length, yes.

Anyhow ... it's simply not necessary .. and this is where opinion based reviews come in handy(such as this Brad Hill fella), and their more extensive findings.

There are many of these Brad Hill's in the world, and in Nikon circles(that I know) the most respected is Bjorn Rorslett, in both his analyses and methods(he gives simple scores) all based on testing in usage not focus targets. He explains if a lens is better at infinity or closer up, and so on and so forth. But his knowledge of technical aspects is pretty much unsurpassed.

Then we may as well not have these test sites at all and just resort to these reports from these users. Personally, I would rather have both, but with the test sites doing a more thorough job, not just a small snippet that only tells part of the story.

Anyway, I am not going to be convinced otherwise as I have more than enough evidence that suggests lenses behave quite differently at distance compared to the standard focus distances that these tests adhere to. You are free to think what you like as I do not want to get into a pi55ing contest.

Just a final note. Here is a link to Lens Rentals and how they test lenses. It might be of note that they suggest that you do a near and far test. Half way down the page on this link:
http://www.lensrentals.com/blog/2010...to-test-a-lens

It states:
"
Near distance widest aperture. Zooms should be checked at their widest, middle, and longest focal length.
If your zoom is variable aperture, be sure to mark down at what distance the aperture changes. You know your f4-5.6 zoom is f4 at the wide end and f5.6 at the long end, but do you know when it changes? Its good to know that shooting at 160mm instead of 170mm will give you f4.
At each location the lens should be checked with aperture wide open, and stopped down in incremental (1/2 or 1 stop) steps to f8. You can assume the lens has reached maximal sharpness by f8. Some may sharpen further in the corners at f11, but they’ll usually start sacrificing center sharpness there.
The same tests should be repeated at far distance.
Testing distances will vary depending on the lens, but I find it logical to test at the distances I’m likely to shoot at. I don’t shoot the 16-35 f2.8 at 4 feet away so near testing may be 10 or 12 feet for it.
Unless you’re going to really shoot at the minimum focusing distance, don’t test there. The lens will usually be sharper a few feet away."

[Sar NOP]

Photozone has tested the 70-200/4.

As expected it's almost as good as the 70-200/2.8VRII ... not better, as some experts seem to think.

It seems that (according to PZ's comments) the VR feature is better than earlier incarnations of VR.
But I'd reckon it has more to do with the lower weight of the lens too.

Why the VR is more efficient on a lighter lens ?
You'd reckon that the VR on a 70-200/2.8 is more efficient than the one on a 200/2 ?

[Lance B]

The Reason the VR is better is because it is the latest incarnation of Nikon's VR, being the new VRIII lens, not a VRII lens. Nikon are touting up to a 5 stop handholdability over the previous VRII which was supposed to be up to 4 stops.

[arthurking83]

I just read Bjorn's review on the VR of this lens, and he says that for all intents and purposes it feels about the same.

The efficiency of VR is always going to be a subjective and personal aspect I guess.

VR on a smaller and lighter lens should allow more handholdability for most users.

That is, at say 200mm with both the f/4 and f/2.8 lens, you should see a greater success rate from the stabilisation system with the f/4 lens at say 1/30s, than you would see with the f/2.8 lens.
This would be more likely due to the lighter weight of the smaller lens.