Sorry to ask: Highest "resolution" lens for D800?

FlowtographyBerlin

Sorry to ask this, because there are tons of articles, reviews blah blah blah on the net, but: Does anyone what is the highest "resolution" lens available for the Nikon mount? By "resolution", I mean LP/mm, which is why I put the word in quotation marks, some people call this "sharpness" :-)

Let's assume we're talking about the center value only and focal length doesn't matter.

I don't want to start a discussion about a best-lens ranking like the DxO list or anything, just get a hint how to find the highest LP/mm lenses available. On DxO Mark, I can't find any numeric values, just color-coded graphs and scores (=ratings), and on i.e. on photozone I can only skip through individual reviews, plus many lenses are tested "only" on the D3x.

Maybe someone can point me to it :-) Thanks!

JJ_SO

Sigma 35/1.4 is a candidate for that: http://www.lensrentals.com/blog/2012/11/sigma-35mm-f1-4-arrives-announces-new-world-order with a Peak wide open of 775 LP/IH at MTF 50

But that looks ridiculous to a Zeiss 70-200/2.9 parfocal with 990 LP/IH at MTF 50. I'm sure these are not the highest possible, just the highest I know of.

At f/2.8, the Sigma shows 1000 LP/IH in the center and might be a bit more affordable than 20k$ for the Zeiss.

Sorry, was talking BS, the numbers are correct, but not LP/mm. Have to reread the MTF article.

After rereading its Linepairs per Image height (LP/IH)

haroldp

At optimum apertures (usually around f5.6 or 6.3), any of the Nikon professional lenses are indistinguishable on a D800e. At those levels, you would have to be at mirror lock and remote release on a very good tripod, and in perfect focus to even think about these differences.

.... H

birdman

I second the other opinions. Look at the Nikon 200/4.0 Micro/Macro and see how much resolution it pulls.
Almost every Micro/Macro, even other manufacturer's such as Tamron 90mm or Sigma 180/2.8 OS, would probably be the sharpest, or some of the sharpest, lenses on Earth -- especially paired with the D800.

The Zeiss 100/2.0.....Zeiss 21/2.8....Nikon 135/2.0......all are ridiculously sharp.

JJ_SO

"sharpness" and "resolution" really are different things. We could add a Zeiss Apo-Sonnar 135/2 to that selection. But saying that I should also mention that the highest resolution might be just not a practical thing, if you need to carry a 20 pound tripod and some windshields around to achieve it.

And Roger is usually testing a batch of the same lens, so one gets also an impression about how much lenses can differ from their colleagues.

Is it just the "guiness book highest resolution" thing or do you have practical reasons to ask, @FlowtographyBerlin ?

Golf007sd

@FlowtographyBerlin: I know you are familiar with the list Nikon put out a few weeks ago that they recommend for the D800/D800E. Thus, your question is to vague. What type of shooting are you planing on doing?

Msmoto

If one looks at the MTF charts...the new Nikon 800mm f/5.6 is almost a straight line across the top.... I believe it becomes more difficult to get full frame sharpness the shorter the focal length do to the issues of retro focus, distortion, light fall off, etc.

But, the issue for me is that almost all the Nikon pro lens will outperform my abilities greatly.

haroldp

But, the issue for me is that almost all the Nikon pro lens will outperform my abilities greatly.

I agree.

Mine as well, particularly hand held, and I think I am very good.

Regards .... H

Ade

If one looks at the MTF charts...the new Nikon 800mm f/5.6 is almost a straight line across the top.... I believe it becomes more difficult to get full frame sharpness the shorter the focal length do to the issues of retro focus, distortion, light fall off, etc.

Unfortunately we cannot determine "resolution" from the kind of MTF charts published by Nikon, Canon, etc. Those charts are meant to show how well contrast and sharpness are maintained across the lens (from center to edge) -- not to indicate resolution.

Nikon MTF charts plot contrast vs. distance from center with the frequencies fixed (at 10 and 30 lines/mm).

What we need instead is a different kind of MTF chart plotting contrast vs. frequency. We can then make some statements about resolution from the maximum frequency where the contrast is at some minimum value (e.g., at 10%).

TaoTeJared

@FlowtographyBerlin: I know you are familiar with the list Nikon put out a few weeks ago that they recommend for the D800/D800E. Thus, your question is to vague. What type of shooting are you planing on doing?

+1

LP/mm or the Lp/Ih for 99% of photography makes little to no difference nor is there much difference between lenses of the same class. (i.e. Nikon pro/ Zeiss, sigma consumer/Nikon consumer, $1,000 vs 1,000, $300 vs $300.) The only time you might be able to tell a difference is in a 100% controlled environments.

MsMoto is correct, any manufacture's macro's are the sharpest - that is what/how they are designed for. Nikon's 60mm resolves the most detail by far in my bag - almost too much for some situations that I have (I have to really work to soften skin in portraits.) The 105vr is next. Probably my 180mm 2.8 is third sharpest. Then my 70-200vr1 but every other lens I own resolves about the same as that to the naked eye - even one's that are 40years old.

FlowtographyBerlin

775 LP/IH at MTF 50
But that looks ridiculous to a Zeiss 70-200/2.9 parfocal with 990 LP/IH at MTF 50. I'm sure these are not the highest possible, just the highest I know of.
At f/2.8, the Sigma shows 1000 LP/IH in the center and might be a bit more affordable than 20k$ for the Zeiss.
Sorry, was talking BS, the numbers are correct, but not LP/mm. Have to reread the MTF article.
After rereading its Linepairs per Image height (LP/IH)

Looks like I also have to do some re-reading, because that would imply that the D800 sensor would outresolve the best lenses (to my knowledge, the D800 sensor has been tested to 1600 lp/ih, was it dpreview?

At optimum apertures (usually around f5.6 or 6.3), any of the Nikon professional lenses are indistinguishable on a D800e. At those levels, you would have to be at mirror lock and remote release on a very good tripod, and in perfect focus to even think about these differences.

Really? I don't know, but I thought I can see a difference between i.e. the 24-120 and a good prime. But maybe that's just an impression.

Look at the Nikon 200/4.0 Micro/Macro and see how much resolution it pulls.
Almost every Micro/Macro, even other manufacturer's such as Tamron 90mm or Sigma 180/2.8 OS, would probably be the sharpest, or some of the sharpest, lenses on Earth -- especially paired with the D800.

Ok, I'll check those, thanks for the hint.

"sharpness" and "resolution" really are different things. We could add a Zeiss Apo-Sonnar 135/2 to that selection.
Is it just the "guiness book highest resolution" thing or do you have practical reasons to ask, @FlowtographyBerlin ?

Yeah, this sharpness vs. resolution thing is a pain. No, it's not about the Guinness book, but yes it's more out of interest. Last week, I was doing some product shooting (yawn), and since some of it involved the surface structure of the products, I was simply starting to wonder what would be the absolutely best choice to get the finest detail.

@FlowtographyBerlin: I know you are familiar with the list Nikon put out a few weeks ago that they recommend for the D800/D800E. Thus, your question is to vague. What type of shooting are you planing on doing?

Yeah, I know the list. But I have a mixed-feeling relationship with DxO Mark scores... After all, those scores are ratings, and at least resolution/sharpness-wise, I can't find where they would publish any measured numbers. If they had a small spreadsheet accompanying the color graphs, that would make the issue a little better.

But, the issue for me is that almost all the Nikon pro lens will outperform my abilities greatly.

Well, that depends on how you define "ability". If you've got the focus right, the subject is not moving substantially, and the camera is steady, then it you should see your lens perform. :-)

What we need instead is a different kind of MTF chart plotting contrast vs. frequency. We can then make some statements about resolution from the maximum frequency where the contrast is at some minimum value (e.g., at 10%).

I think at least Nikon actually measures the lenses. I heard the Canon charts are theoretical values from their calculations.

MsMoto is correct, any manufacture's macro's are the sharpest - that is what/how they are designed for. Nikon's 60mm resolves the most detail by far in my bag - almost too much for some situations that I have (I have to really work to soften skin in portraits.) The 105vr is next. Probably my 180mm 2.8 is third sharpest. Then my 70-200vr1 but every other lens I own resolves about the same as that to the naked eye - even one's that are 40years old.

Ok, thanks for the tips!

Thanks to everyone for all your reponses so far!

Msmoto

Here are the lenses I would not hesitate to use on a D800.....but remember I have only a D4 to judge by....
10.5mm f/2.8 (modified hood), 16-35mm f/4 (but has distortion), 24mm f/1.4, 24mm f/3.5 PC, Sigma 35mm f/1.4, 85mm f/1.8, 105mm f/2.8 VR micro, 135mm f/2, 70-200mm f/2.8 VRII, 400mm f/2.8, all Nikkor except the one Sigma.

Lenses I might hesitate are 24-120 VR f/4, 50mm f/1.4D, but these may do well in some situations.

Not all are super sharp wide open, but this is usually the case.

JJ_SO

I read about the 24-120, it has it’s real strength between 35 and 70mm. Ah we'll, the 14-24 has also very good resolution. And the Zeiss 21 and 15mm can also join the row.

TaoTeJared

Lenses I might hesitate are 24-120 VR f/4, 50mm f/1.4D, but these may do well in some situations.

I think the 24-120 VR f/4 works great if you use it for what that lens is built for. Others resolve more, but it is much better than any of the consumer lenses I have tried.

50mm f/1.4D has the same issues as most AFD lenses, CAs, flair and reflective ghosting - It is really not that bad but both the new Gs are better. I have some time this week, maybe I can set something up with all the 50's and 50 equivalents I own.

I still look at all of the "resolution" talk as idiotic - unless you are doing something like product shots or the like. Each lens is designed for a specific use or set of uses and "resolution"/"sharpness" (however people mash those together) is different on every lens. For instance on 1.4s, I want the corners to be softer - that means the bokeh will be better and you have better subject separation. Add to that, when you raise the ISO the resolution drops as well - then it really doesn't make any difference.

Golf007sd

Each lens is designed for a specific use or set of uses and "resolution"/"sharpness" (however people mash those together) is different on every lens.

+1

A few days ago I read this article. It does a great job in educating a person on how to better understand an MTF chart and how we should look at lens in respect to: Contrast and Resolution, Perceived Sharpness, Sharpness: Resolution and Acutance.

birdman

+2 for TTJ. The 24-120/4.0VR is a real "keeper" as Golf007 told me once. It's got stronger FLs, for instance the middle 50, 85mm FLs have shown to be sharper, or perform better, than say, 120mm. Same applies for ALL ZOOMS. None of them work equally the same across all FLs.

Primes are typically sharper and especially so for macros. They're optimized for corner to corner sharper and maximum DOF to see detail. The trade-off has historically been unreasonably slow AF speed. But that's changed as well, as most new macros focus quite fast.

Ade

MsMoto is correct, any manufacture's macro's are the sharpest - that is what/how they are designed for.

Sorry TTJ (and MsMoto), but the above statement is not correct, and is not supported by data. Macro lenses typically perform better than regular lenses at macro working distances. But that doesn't mean macro lenses have higher maximum resolution than regular lenses, at any distance. And in fact, macro lenses often have below-average performance when used at non-macro working distances.

While it is true that macro lenses generally have excellent resolution, they are specialized lenses with some constraints:

- As mentioned, macros are optimized to work at very close distances
- Due to DoF considerations, they are designed to work well at very small apertures
- Being specialized, macro lenses are not always 'refreshed' with the latest technology

Take Nikon's acclaimed 105/2.8 VR, for example. At non-macro working distances, the resolution of this lens -- while still excellent -- is actually below average when compared to other pro Nikkor lenses.

Only at macro working distances does the 105's resolution start to approach the other pro Nikkors, but even then, I believe the 105's measured max resolution is still lower than any of Nikon's trinity zooms (14-24, 24-70, 70-200).

Golf007sd

@Ade: The 105 2.8 Micro VR II is a Pro-Nikkor lens. Given that it is also a prim lens, its performance is fantastic. I would put its sharpness, in addition to resolution (at any f-stop) right there with any of the Trinity set. I own each of these lenses and have seen their performance first hand.

Some sample images for you to judge for yourself. All images taken with a D4.

Nikon 105 2.8



Nikon 14-24 2.8 @ 24MM



Nikon 24-70 2.8 @ 70mm



Nikon 70-200 @ 200mm

Ade

Um, nowhere did I say the 105 VR is not a pro lens. It's a lens I almost always have in my camera bag.

Fact: the 105 VR's resolution, when measured, is below average when compared to other Nikon pro lenses at non-macro working distances.

Fact: the 105 VR's max resolution at macro distances, when measured, is less than (or at best equal to) the max resolutions of the trinity zooms (at non-macro distances).

Please see just about any Nikon lens test database of your choice, and you will see that the above facts are correct.

So the statement and often repeated myth that any manufacturer's macro lenses "are the sharpest" is not correct.

Feel free to post any measured data to the contrary.

Golf007sd

@Ade: I'm pleased to hear you have the 105 2.8 and are enjoying it in a manner you feels is best. It just seems that they way you where describing it, it came across as not being a "pro" lens.

With respect to your position that the 105, or for that matter any macro lens, lack the resolution when compared to other Nikon pro lenses, would you be so kind to proved me (us) with some insight in how you have come to this conclusion. Any article, data you can supply me (us) would be greatly appreciated. Thank you.

Ade

As mentioned, just about any reputable lens test database would show the above to be correct.

Here's the data from the folks at lensrentals.com, all tested recently using D800 on an Imatest rig:

105/2.8 VR resolution (normal distance) center: 934 average: 870
105/2.8 VR resolution (macro distance) center: 1050 average: 990

14-24/2.8 resolution (normal distance) center: 1181 average: 831
24-70/2.8 resolution (normal distance) center: 1107 average: 902
70-200/2.8 resolution (normal distance) center: 1080 average: 957

All measurements in LP/IH.

While we're at it, I noticed that they even wrote a blog entry noting the 105 VR's average performance at normal (non-macro) focus distances:

http://www.lensrentals.com/blog/2012/07/imatest-macro-results-with-apologies-to-the-nikon-105-vr-micro

Paperman

As mentioned, just about any reputable lens test database would show the above to be correct.

Here's the data from the folks at lensrentals.com, all tested recently using D800 on an Imatest rig:

105/2.8 VR resolution (normal distance) center: 934 average: 870
105/2.8 VR resolution (macro distance) center: 1050 average: 990

14-24/2.8 resolution (normal distance) center: 1181 average: 831
24-70/2.8 resolution (normal distance) center: 1107 average: 902
70-200/2.8 resolution (normal distance) center: 1080 average: 957

All measurements in LP/IH.

While we're at it, I noticed that they even wrote a blog entry noting the 105 VR's average performance at normal (non-macro) focus distances:

I'm a bit confused here. " Good " lenses are usually around 3500 lw/ph center at MTF50 tests - 3000 lw'ish for wide open . Divide that by 2 for pairs , you are between 1500-1750 lp/ph . ( Photozone/Dpreview )

Where does lensrentals get their scores from ? Are you sure they are MTF50 and not for some higher contrast ? Or do they have a different image/picture height ??

FlowtographyBerlin

Lenses I might hesitate are 24-120 VR f/4, 50mm f/1.4D, but these may do well in some situations.

I think the 24-120 VR f/4 works great if you use it for what that lens is built for. Others resolve more, but it is much better than any of the consumer lenses I have tried.
[…]
I still look at all of the "resolution" talk as idiotic - unless you are doing something like product shots or the like. Each lens is designed for a specific use or set of uses and "resolution"/"sharpness" (however people mash those together) is different on every lens. For instance on 1.4s, I want the corners to be softer - that means the bokeh will be better and you have better subject separation. Add to that, when you raise the ISO the resolution drops as well - then it really doesn't make any difference.

Of course, you're right on this, TTJ, but this is exactly the kind of discussion I wanted to avoid. My question wasn't about great lenses, but really simply about the highest resolution/contrast performers. And it is exactly the product shot case where this question came up. Studio tripod, good lighting, low ISO.

The fact that there is now zooms being defended that they have FLs with good performance etc. is also more of the "good lens" discussion category and not on the criterion I was asking for.

Yes, the zooms are probably good for a zoom, but I would doubt that any (excellent) zoom lens would be able to compete with an excellent prime in the matter. Yes, any of the lenses that anyone with a little experience and feel for the matter uses have their use cases and their specific character that makes them so great for different uses, and yes sharpness, resolution, contrast are only one bunch out of the many criteria that one would use to describe lens performance.

In this case, though, all I was focussing on was the mere isolated criterion of high contrast/high resolution.

Thinking about it, my FL criterion has to be slimmed down to higher numbers, though: I rarely use wide-angle in product shots...

The info on the macro lenses is good, I'll check those out. I actually haven't ever used any of those. So the compiled list from your hints is so far:

Nikon 60mm Micro
Zeiss 100/2.0
Nikon 105 VR
Nikon 135/2.0
Zeiss Apo-Sonnar 135/2
Nikon 200/4.0 Micro

Next time I have a paid product shoot, I'll rent one or two of these and check'em out. Too bad my rental place has Zeiss only for Canons.... And no 200/4 Micro.... Hm... Will have to check other rental places.

Thanks for all your input, everyone!

FlowtographyBerlin

Uh, more comments while I typed. Sorry.

@Pitchblack: Yes, but I was referring to this already in what I said in my last answer.

70-200/2.8 resolution (normal distance) center: 1080 average: 957
All measurements in LP/IH.

I'm a bit confused here. " Good " lenses are usually around 3500 lw/ph center at MTF50 tests - 3000 lw'ish for wide open . Divide that by 2 for pairs , you are between 1500-1750 lp/ph . ( Photozone/Dpreview )

Are you sure they are MTF50 and not for some higher contrast ? Or do they have a different image/picture height ??

+1: I'm confused, too, as I already said above with the other Lensrentals data that @JJ_SO posted. From what I read, a sensor typically/roughly resolves about a two thirds of its pixel resolution, divide that by two and that makes it something a little above 1600 LP/IH for the D800 (4912 px height). I can't imagine that the most excellent prime lenses are outperformed by the D800 sensor.

One thing is that the tests i.e. on the Zeiss 135 were made on a Canon 5DII (3744 px height), but that should still leave a sensor resolution of approx. 1200 LP/IH.

Ade

MTF test results are not comparable between different setups / systems / methods. For example, much of the higher LP/IH numbers were probably obtained from testing JPEG images, whereas lensrental analyze the RAW files. Using different test charts, lighting conditions, etc., will all also produce different results.

The benefit of the lensrental data is that all of those MTF numbers were obtained using the same setup & methodology on the same camera (D800 in this case).

Also see:

http://www.lensrentals.com/blog/2012/09/why-arent-the-damn-numbers-the-same

Ade

Yes, the zooms are probably good for a zoom, but I would doubt that any (excellent) zoom lens would be able to compete with an excellent prime in the matter.

This is another myth. Please look again at the MTF data presented earlier comparing the 105 VR (prime) vs the three trinity zooms. In terms of resolution, it is not unusual for zooms to equal or exceed primes.