Lens suggestions...

I find the digital picture com’s tests poorly done, it is obvious the shots with some Nikon lenses are not correctly focused.

 

I think the problem is that people don’t understand that lens design is a compromise and you cannot expect improvement in one area not to cause some losses elsewhere. Personally I think people pay too much attention to MTF tests and too little to how the images actually look to the human eye.

You're absolutely right about the overall look and what not up until you run into some "technical" photography, especially in the commercial fields, where you're expected to deliver images that are as sharp as possible in each and every area of the frame. Your guy may like the "look" of the lens or how the zoom feels but if you read my posts you'll see that I did say that a lens is so much more than its potential resolving power. However, this is not what it's about. I don't give a rat's ass about what some guy says about a lens until I see the PICTURES. And that's exactly what Bryan does with his site.. You say that the pictures can't be right? I beg to differ because I've tested some of those lenses myself and saw the same things. And BTW, he uses Live View to focus and he does those tests for a living. You're really gonna say that the focus must be off?

Thedigitalpicture is NOT popular with the Nikon die-hards because often the truth is hard to swallow. In some ways It's like giving up a religion. And as for me... Well, let's just say that one morning I woke up and smelled the roses and I ain't never looking back.

But if the love for your Nikkor lenses is strong - great! Keep enjoying them. Each and every Nikon glass that you have - there's not another one like it. So, if you love the "look" - stick to it. I was talking only about image sharpness, and trust me, those pictures are dead-on.

To each is own, I guess, and it's so true.

Thanks, David. I'm not in general going to comment on the site's reviews in general, and I'm under no delusion about the perfection of any lenses.

Nonetheless, the centre of the frame at 200mm should be where the new 70-200 E is strongest, and no analysis I've seen has reported it being substantially softer than at, say, 135mm; I'd not raise more than an eyebrow if the corners alone had been off. Yet that particular image set looks troubling, where the other focal lengths seemed less so. I, too, value actual images (and miss that dpreview stopped doing them) - although there's always a danger in reading too much into the performance at a single focal length.

Mistakes can be made when reviewing (as I've seen on other sites), and equally there can be problems with individual lenses that are tested (Roger Cicala at LensRentals does an interesting job of graphing variation between a brand). This particular image set looks either like an error or a lens fault, and I'd like separate confirmation that this is a predictor of the behaviour of that lens.

Are you saying you've tried the 70-200 E series and found that the 200mm performance is a problem? I'd like to know, if you're offering a second data point. That lens is currently on my "to purchase" list, so I have a vested interest in determining whether it's actually going to have a problem at a highly-used focal length - and the reason I'm considering an upgrade from my mk2 is precisely because I do want better performance at f/2.8.

Thanks, David. I'm not in general going to comment on the site's reviews in general, and I'm under no delusion about the perfection of any lenses.

 

Nonetheless, the centre of the frame at 200mm should be where the new 70-200 E is strongest, and no analysis I've seen has reported it being substantially softer than at, say, 135mm; I'd not raise more than an eyebrow if the corners alone had been off. Yet that particular image set looks troubling, where the other focal lengths seemed less so. I, too, value actual images (and miss that dpreview stopped doing them) - although there's always a danger in reading too much into the performance at a single focal length.

 

Mistakes can be made when reviewing (as I've seen on other sites), and equally there can be problems with individual lenses that are tested (Roger Cicala at LensRentals does an interesting job of graphing variation between a brand). This particular image set looks either like an error or a lens fault, and I'd like separate confirmation that this is a predictor of the behaviour of that lens.

 

Are you saying you've tried the 70-200 E series and found that the 200mm performance is a problem? I'd like to know, if you're offering a second data point. That lens is currently on my "to purchase" list, so I have a vested interest in determining whether it's actually going to have a problem at a highly-used focal length - and the reason I'm considering an upgrade from my mk2 is precisely because I do want better performance at f/2.8.

At first I didn't really understand what you meant but on a second, very long look and after squinting very hard I think that now I do. Because of the fairly significant vignetting, which you would normally get at this kind of setting, the parts of the frame that are further away from the center have registered less light and some of the blacks have dropped below the threshold. In the center, however, all of the blacks are above the zero. So, the mid-frame and corner samples display deeper blacks. What that causes is that the more gradual transition which occurs in reality, here, appears to be happening more abruptly, because the intermediate tonal values are clipped. The ever-so-slightly darker edges give the illusion of an increased sharpness and the slightly deeper blacks impart an overall feel of slightly more contrast. Also, the absence of the same fine lines in all the sample but the center one does not help the issue. The very fine lines in the center appear more blurry by comparison, than the thicker ones in the other samples. It is an illusion. I had no problem with these very samples before and I don't have one.

And no, I did not get to run my tests on the new 70-200 FL Nikkor. Regardless, all the test samples from the familiar to me lenses have been in line with my own tests / experience so far.

In regards to any lens's sharpness this has been my go-to site for a long time and I haven't been let down yet. However, I do realize that there are variations in the production as well as "lemons" and a number of below-spec copies of lenses, apparently, is featured at thedigitalpicture.com. So, the test samples that you see there should be taken with a grain of salt. And if there is a similar web site that lets you compare test shots in a comprehensive manner from a large selection of concurrent photographic lenses, please take me to it. My point being this is as good as it gets and we have to start somewhere, I know I do. You yourself implied that MTF graphs don't always do justice to a lens, so what is better for assessing any given lens's sharpness other than the actual photos, especially ones taken under test conditions?

And, Andrew, just so you know, Bryan does sometimes repeat his tests if there's a reason to believe that there's a problem with the samples he already has on file. Samples from more than 1 copy of a lens are available on select lenses. It's usually 2 but I've seen even 3 variations on a couple of lenses and maybe even 4 on one. If there are samples from an additional copy, you will see a small drop-down menu right next to where you select the camera body. Here's a comparison between 2 different copies of the Tamron 24-70mm f/2.8 Di VC USD G2:

Tamron 24-70mm f/2.8 Di VC USD G2 Lens Image Quality

It would be very unusual for the edge to be sharper than the center. I am fairly convinced it is a focusing issue. LV doesn’t guarantee best focus especially in an indoor environment there may not be enough light. Photozone focus bracket their shots which is probably why I’ve found better consistency in their results vs. my own experience. I normally shoot my images of people with autofocus and have the focus point on the subject in an off center position. If one focuses on the center, the off center areas may be off because of alignment of the test, the lens, or field curvature. For this reason I prefer tests that focus bracket and report the best result or at least focus in each part of the test target separately so the best focus is obtained for each area.

I think perhaps the problem is the consistency of Nikon’s LVAF (this is certainly an issue with the D3X; indoors one would get quite erratic LV focus with that camera; newer Nikons have improved LV image quality as well as LVAF).

With regards to lens tests I think each site have their own strengths and weaknesses. Lensrentals test only wide open at infinity focus but they test many samples. This gives them a great deal of credibility in my eyes, but for lenses that excel at shorter distances it may give a misleadingly poor result. Lenstip use six test distances and present the average, if I recall correctly. Dxomark test using multiple different camera bodies so you can see the effect of sensor on resolution. Photographylife’s Mansurov uses flash for his long lens tests which takes out the vibration of the lens mount out of the equation. Lenscore use a 200MP sensor for all tests so you can predict performance with a future high resolution sensor (and get results that tell about the lens using essentially an ideal sensor). I think the best approach to viewing lens test data is to use multiple sources. Otherwise we do not get a complete understanding of the lens performance characteristics.

But, in the end one’s own practical use case may be something that doesn’t match lens test site conditions and ultimately viewing one’s own images tells the decisive story. Still, if one is not happy with some aspect to a lens’s performance, lens test sites can sometimes provide a clue as to why the results are the way they are. But this requires a good understanding of the test procedure and differences to how one uses the equipment in real world situations.

I have personally been very happy with most of Nikon’s latest lenses (such as the 1,4’s and 2.8 zooms) but I admit that the prices have gotten out of hand.

It would be very unusual for the edge to be sharper than the center. I am fairly convinced it is a focusing issue. LV doesn’t guarantee best focus especially in an indoor environment there may not be enough light. Photozone focus bracket their shots which is probably why I’ve found better consistency in their results vs. my own experience. I normally shoot my images of people with autofocus and have the focus point on the subject in an off center position. If one focuses on the center, the off center areas may be off because of alignment of the test, the lens, or field curvature. For this reason I prefer tests that focus bracket and report the best result or at least focus in each part of the test target separately so the best focus is obtained for each area.

 

I think perhaps the problem is the consistency of Nikon’s LVAF (this is certainly an issue with the D3X; indoors one would get quite erratic LV focus with that camera; newer Nikons have improved LV image quality as well as LVAF).

 

With regards to lens tests I think each site have their own strengths and weaknesses. Lensrentals test only wide open at infinity focus but they test many samples. This gives them a great deal of credibility in my eyes, but for lenses that excel at shorter distances it may give a misleadingly poor result. Lenstip use six test distances and present the average, if I recall correctly. Dxomark test using multiple different camera bodies so you can see the effect of sensor on resolution. Photographylife’s Mansurov uses flash for his long lens tests which takes out the vibration of the lens mount out of the equation. Lenscore use a 200MP sensor for all tests so you can predict performance with a future high resolution sensor (and get results that tell about the lens using essentially an ideal sensor). I think the best approach to viewing lens test data is to use multiple sources. Otherwise we do not get a complete understanding of the lens performance characteristics.

 

But, in the end one’s own practical use case may be something that doesn’t match lens test site conditions and ultimately viewing one’s own images tells the decisive story. Still, if one is not happy with some aspect to a lens’s performance, lens test sites can sometimes provide a clue as to why the results are the way they are. But this requires a good understanding of the test procedure and differences to how one uses the equipment in real world situations.

 

I have personally been very happy with most of Nikon’s latest lenses (such as the 1,4’s and 2.8 zooms) but I admit that the prices have gotten out of hand.

I totally agree on the multiple data point thing. Also, I wanted to but forgot to mention that I do realize that the field curvature is a factor in Bryan's tests because of the relatively short focusing distances. I have run into this issue in my own tests on multiple occasions. So, quite a few lenses will have more corner sharpness out in the field than what can be gathered from Bryan's test samples. Personally, I concentrate on the ~f/5.6 to f/16 range in my own lens tests but even there I've seen some weird stuff like: corners that are sharper at f/8 than f/5.6, while the center being slightly less sharp due to diffraction. Having that happen in only one half of the frame would easily explain it as the image plane not being parallel to the target but having it the same in all 4 corners? That's field curvature to ya.

I still do not believe that the focus accuracy is an issue in any of the test samples on thedigitalpicture. One can focus with the LV very accurately, even if the mechanism itself is acting up. In my own tests I direct a bright beam from an LED flashlight that's mounted to a light stand onto where I'll be focusing. Then I zoom in to 100% in LV and engage the AF a few times to "get the feel" for the sharpest obtainable image. After I'm satisfied with the focus I then often fiddle with the focus ring to make sure that I didn't miss a point of an even more accurate focus. After that I may engage the AF again or simply return the focus manually to the "in focus" point, by visual inspection. I do that when I sense that there may be a problem with the consistency of the AF. I am sure that Bryan does his best to keep any focusing errors within a very small margin. * I use a flash to capture the test image itself, obviously and I like the fact that someone out there actually cares about the vibrations introduced by the shutter. There may be some benefit in popping the flash several seconds after the shutter has opened. I don't believe I used the technique in my own tests, however, I used it quite a lot when I was doing macro work in a studio. That was back in the day when I was shooting with my F5. Locking up the mirror was not very viable because it would, potentially, offset the composition and/or focus, so, I would turn off all the lights, open up the shutter in Bulb and manually set off a few bursts (the required amount), because the lens was often stopped down to f/16 or f/22. That would take care of any unwanted movement. Also, I would use my "flashlight" technique to determine the desired DOF and the optimal focus point. Live View is a godsend, I can tell you that much.

If talking only about sharpness, then all those lens tests are just fine and dandy. But in the end, it's about photos.

When talking actual images, there are just too many mindblowingly good lenses that test mediocre or poorly. All those tests on 2D objects fail to reveal the complete truth, apparently. I read a fair amount of lens tests, but always look for images first and foremost. And a lot of those lenses that test great for sharpness, corner performance and all that - the photos just do not convince me. It almost seems that some lenses achieve a level of sharpness and high contrast that has become too biting, yielding images that look coarse and almost unelegant. I really don't care for the look, and often enough feel it's actually degrading.

And then there are those lenses with other design briefs - like the 105 f/1.4E, the 58 f/1.4G, several Zeiss lenses - that maybe have less impressive charts but once confronted with an actual 3D object, things just work. Maybe not measurably, maybe not for all kinds of photography - but great tools that deliver that bit extra that is not achieved with those lenses that test perfectly in the lab. And those who feel those lenses with great lab results are superior will probably feel the lens is degrading the image.

Nobody is right or wrong here, but we should admit there is not one single holy grail in this discussion, and hence nor a single way of approaching lens tests that will give conclusive results for all.

So, rather than thinking that Nikon stops making lenses, I hope they (dare to) make more lenses that are designed to be different, and do not go down the rabbit hole of achieving lab greatness like many others seem to do. This is *not* about "strong love for your Nikkor lenses", but rather a case of horses for courses. I don't really care about who produces the lens, but I do care about other things that perfect flatness of field, corner sharpness and the perfect MFT graph.

Which is a long way of saying: I don't trust choosing lenses over results in test labs. Not because the tests are potentially flawed (or not), but because these tests are limited by design, and do not and cannot tell the whole story.

If talking only about sharpness, then all those lens tests are just fine and dandy. But in the end, it's about photos.

 

So, rather than thinking that Nikon stops making lenses, I hope they (dare to) make more lenses that are designed to be different, and do not go down the rabbit hole of achieving lab greatness like many others seem to do.

Yeah, so, stop making them in China while charging $$$$ over the competitor's, then we'll talk business. Personally, I don't believe that I will ever buy another Nikkor lens again, unless Nikon makes something that blows me away in regards to image sharpness while doing something unfathomable about the price tag.... So, nope... not holding my breath.

So what you're saying is that Chinese people shouldn't be paid proper wages, or did I understand you incorrectly? I think that for the same job equal wages should be paid irrespective of location (and so products would roughly cost the same across regions of manufacture unless one region has some advantages in infrastructure, education, material availability etc.). Nikon may have factories in China partly because there is limited space and resources on the Japanese islands and it makes sense to take advantage of the resources of China (and Thailand etc.) not just for lower costs (which may not be quite so low any more) but also for availability of educated personnel and space.

I think image sharpness is overrated (there is an abundance of detail beyond practical needs with almost any modern lens and camera). I would emphasize other characteristics of a lens now (since sharpness is so easy to get beyond what can be communicated to others through realistic means). I consider how the image is rendered in backlight, rendering of skin, out of focus areas, autofocus accuracy, colour, how the lens handles etc. equally important. And the overall feeling of images rendered by the lens above all. Most of my images are not printed at very large sizes and inspected with a loupe and so I'm interested in image quality which can be communicated in any size.

While one of the things that makes me cringe about KR's web site is his tendency to write "Quality: Made in China" on his reviews, I don't think it's reasonable to assume that everyone will get paid the same globally. I don't assume that things made in China (by a reputable company) necessarily have quality concerns (I believe Nikon make premium items in Japan because for small runs it's better to be near the R&D centre, whereas ramping up bulk items in a factory works better in China, but they really have to ensure the mass-produced items work properly, and it's been the high-end Nikons which have had recalls). I'm absolutely aware that there are some human rights issues (I'm British, I have several centuries of not having the moral high ground on this) and (especially historically) intellectual property issues in some regions; China is absolutely not alone in this.

But it's also true that the cost of living is very different in different locations; even in the UK, I could double my salary by working for a company in the centre of London, but it would cost me a large chunk of that to commute or live closer in. Some locations are necessarily competitive because a given salary is "a good wage" relative to the living costs of the area, but impractical in other countries. Eventually this balances out - I believe there are locations in India, for example, where so many IT companies are clustered and competing that the employee salaries are approaching those in the US. Nevertheless, you would expect a worker of moderate qualification in a Chinese factory to be paid less than someone working in London or Silicon Valley doing the same job - there is much more competition for the same job. Additionally, there's more space - the factory itself in China is likely to be much cheaper due to competition for land.

So are Nikon charging a lot of money for a lens compared with its production costs? Probably. But they still need to ensure their reputation for quality control - there have been relatively few truly awful Nikkors. They can't share production across mounts in the way that Sigma can, increasing their comparative running costs. And yes, they charge more because they know they can, and people will still buy them. You can't really criticise a company for doing that - they exist to make profit.

Sigma, to my mind, are doing absolutely the right thing to deal with that - the "Art" line have very good reputations (more so than the old "EX" branding), and people are prepared to pay a premium for them despite the company, as I did with the 50mm Art. People are starting to hope for an "Art" version of their individual lenses. Tamron are beginning to make similar inroads. Nikon are, arguably, responding with competitively-priced lenses like the 200-500 - it's the premium for some legacy primes that suddenly look unreasonable, but they were completely justifiable when the 85mm f/1.4 AF-S and 50mm f/1.4 AF-S actually were dominating the test charts.

I suspect both Tamron and Sigma have realised that it's not economically viable to subsist on budget superzooms. Nikon, ironically, still make many 18-xxx variations, although a lot get bundled with cameras. As the third parties make technological advances and can justify living with the pricier Nikons, they increase their profit margins, but Nikon are caught justifying their reputation. As they respin lenses they may well be able to do so, but they have a big back catalogue to get to, and Nikon have struggled to get many lenses out the door recently; whether that's cost cutting, issues in Japan or a focus on format changes is TBD.