Test, glass Good enough for D850 !

Important, this writing will contain hyperbole's and big words that is only related to a image that is about 6,2ft wide. If one only see whole pictures at computers or small prints as A2 or A1 this writing is not very relevant and the reader should skip this thread unless pixel peeping is of interest. I usually print 40-63" wide pictures so this have a relevance for my printings and i just share here my findings.

My Conclusion is that i am unlikely to buy another high mp FX camera as lenses stands today. Even if a "D5s 60mp 20fps with hybrid viewfinder" comes along. Because few lenses will keep up with a 60mp sensor. Another point is that the prices of lenses for FX have become so high that i am considering a MF for landscape, if one look at Fuji GFX today the lenses are at the same cost as CaNikon, but the sensor is 68% larger and 2x the mp and vastly better noise performance when the 100mp comes next year.

I did not include my Tamron 24-70 Zooms, i have both G1 &G2, but they are not even close to be included, they are downright unsharp and i don't think they even yield half the mp of the sensor. (but remember there are no sharp 24-70 in the market at all)

In my Nikon 70-200/f2.8 FL review i posted comparisons to the Tamron 70-200 VC G1, and the Nikon is clearly sharper, and here compared to the 14 year old Nikon 200/f2 the 70-200/f2.8 FL lens is beaten, but how good glass do one need to utilize fully a D850 ?



left, Nikon 85 TS, center of frame.





Left side





Nikon`s, in lower left corner, the 200/f2 is waaay sharper than the 70-200 FL



Center of frame.



@f5,6 both Nikons are close to perfect even in top right corner, they are equal here.



Tamron does not frame 200mm, here compared to the Nikon 70-200/f2.8E FL





A favorable spot for comparing the Tamron as it have some uneven sharpness in the frame.



But it generally looks like this against the Zeiss 135/f2





Honorable mentions, Sig 50A and Nik 400/f2.8 is also sharp, almost as Zeiss. If one is using Tamron zooms or Nikon 70-200/f2.8 VR2 (or Nikon 24-70 zooms) one is better to stick to the older 36mp sensors until one have updated to the new Nikon 70-200FL lens, and/or other good primes, as one can see over here anything but the very best is diminishing returns for those who upgrade from D800`s to a D850. Money will absolutely yield more if one bought better glass on a D800/810 than to only upgrade to an D850 without getting/having the very best lenses that exist.

Well, in my opinion, even if Nikon produce a D5s ( which they probably will in the future), it will have the same mega pixel capacity as a D5.

You would be looking at a D900 or something like that . Who knows what is in Nikon's planning?


Regards Peter

I'm still trying to figure out why you want to print 60" images of brick walls . . . ;-)

I'm also confused as to whether we are talking about sharpness or resolution . . .

north w said:

Important, this writing will contain hyperbole's and big words that is only related to a image that is about 6,2ft wide. If one only see whole pictures at computers or small prints as A2 or A1 this writing is not very relevant and the reader should skip this thread unless pixel peeping is of interest. I usually print 40-63" wide pictures so this have a relevance for my printings and i just share here my findings.

My Conclusion is that i am unlikely to buy another high mp FX camera as lenses stands today. Even if a "D5s 60mp 20fps with hybrid viewfinder" comes along. Because few lenses will keep up with a 60mp sensor. Another point is that the prices of lenses for FX have become so high that i am considering a MF for landscape, if one look at Fuji GFX today the lenses are at the same cost as CaNikon, but the sensor is 68% larger and 2x the mp and vastly better noise performance when the 100mp comes next year.

I did not include my Tamron 24-70 Zooms, i have both G1 &G2, but they are not even close to be included, they are downright unsharp and i don't think they even yield half the mp of the sensor. (but remember there are no sharp 24-70 in the market at all)

In my Nikon 70-200/f2.8 FL review i posted comparisons to the Tamron 70-200 VC G1, and the Nikon is clearly sharper, and here compared to the 14 year old Nikon 200/f2 the 70-200/f2.8 FL lens is beaten, but how good glass do one need to utilize fully a D850 ?

Click to expand...

A 501.8D at f8 would work fine.

Virtual Photon said:

I'm still trying to figure out why you want to print 60" images of brick walls . . . ;-)

Click to expand...

The wall have always a big print hanging in it, i just took down a huge (86" diagonal) frame that i handmade as a prototype with 140x44mm wood profiles.

On the brick wall is usually a 40x63" print size with 90x33mm frame profiles of beech or Oak. Now there is hanging a 60" print taken in Oregon.

Last week i got the D850, but work and traveling got in the way so i have only used it once outside. And weather have been on/off rain/snow sideways at 30+mph so a equipment test of the brick wall was a nice way to get to use the D850. I have not tested these lenses in a light controlled environment so that i actually have seen the exact difference between the lenses, so now i know their strength and weakness, (the test only confirmed my experience)

I wanted to know that my brand new Tamron 24-70 and Nikon 70-200FL was ok before i take the 3week trip abroad in mid December to Wa&Or to photograph the area again. It turned out that the Tamron 24-70 was significant de-sentered, it was really good and sharp on the left side and in center, but on the outer right side, meaning 1/4 of the picture it was fuzzy, so the dealer is sending me a new copy but it might not get in time before the trip, (he had also 2 other with the same problem).

Virtual Photon said:

I'm also confused as to whether we are talking about sharpness or resolution . . .

Click to expand...

Only sharpness is of interest. MP is fixed here at 45Mp, but Mp don't tell much of sharpness, bacause junk glass inn= junk file out, and visa versa.

Mp in this context have a lot less to do with sharpness than lenses. Regarding sharpness one should first buy the lenses,.. than the D850, keep the D800/E/810 until one have bought all the best glass, 36/45mp is a minute difference compared to lenses.

As one can see above, barley a handful of lenses can keep up below 200mm on a D850 so the pixel war is stopped by lens sharpness, diffraction and MF-prices on FX lenses.

Rounding up as Captain Obvious here. Better sharpness from now on is easier/better to achieve on MF. A larger sensor/target for the light to hit=sharpness. I assume Fuji and Hasselblad will multiply sales with the 100mp sensor. X1D is very tempting, i know they have the glass for it, (i have X1D files). With fixes on the SW and maybe some other upgrades i think MF will sell if the price trend on MF is continuing..

Reilly Diefenbach said:

north w said:

In my Nikon 70-200/f2.8 FL review i posted comparisons to the Tamron 70-200 VC G1, and the Nikon is clearly sharper, and here compared to the 14 year old Nikon 200/f2 the 70-200/f2.8 FL lens is beaten, but how good glass do one need to utilize fully a D850 ?

Click to expand...

A 501.8D at f8 would work fine.

Click to expand...

I am confident it wont. I had the Nikon 50/f1.4 that is equal sharp to the f1.8 version according to DXO, and my Sig 50A is significant more Sharp on my D800, this would only get magnified on the D850, so the 50/f1.8 will far from utilize a D850 and is a mismatch on a high mp sensor.

north w said:

Reilly Diefenbach said:

north w said:

In my Nikon 70-200/f2.8 FL review i posted comparisons to the Tamron 70-200 VC G1, and the Nikon is clearly sharper, and here compared to the 14 year old Nikon 200/f2 the 70-200/f2.8 FL lens is beaten, but how good glass do one need to utilize fully a D850 ?

Click to expand...

A 501.8D at f8 would work fine.

Click to expand...

I am confident it wont. I had the Nikon 50/f1.4 that is equal sharp to the f1.8 version according to DXO, and my Sig 50A is significant more Sharp on my D800, this would only get magnified on the D850, so the 50/f1.8 will far from utilize a D850 and is a mismatch on a high mp sensor.

Click to expand...

No, it most certainly is not. The 50D will resolve all available pixels center to mid at f8, and will simply destroy any zoom lens. The corners are a bit better on the Milvus and the Sigma, but that's about it. Internal reflections/veiling flare are better handled by the higher end lenses.

Reilly Diefenbach said:

north w said:

Reilly Diefenbach said:

north w said:

In my Nikon 70-200/f2.8 FL review i posted comparisons to the Tamron 70-200 VC G1, and the Nikon is clearly sharper, and here compared to the 14 year old Nikon 200/f2 the 70-200/f2.8 FL lens is beaten, but how good glass do one need to utilize fully a D850 ?

Click to expand...

A 501.8D at f8 would work fine.

Click to expand...

I am confident it wont. I had the Nikon 50/f1.4 that is equal sharp to the f1.8 version according to DXO, and my Sig 50A is significant more Sharp on my D800, this would only get magnified on the D850, so the 50/f1.8 will far from utilize a D850 and is a mismatch on a high mp sensor.

Click to expand...

No, it most certainly is not. The 50D will resolve all available pixels center to mid at f8, and will simply destroy any zoom lens. The corners are a bit better on the Milvus and the Sigma, but that's about it. Internal reflections/veiling flare are better handled by the higher end lenses.

Click to expand...

ehhh, did you not see or read the last attachment from DXO i pasted, the 50/f1,8D is rated quite low by today's standards, and do not even resolve the center of a D800e at @f8. On a D850 it will only get magnified further vs Sigma Art, this is flagrantly obvious. DXO in general rate 1.8D as 24mp and the Sigma art as 35mp and i think that is quite correct.

And the 1.8D is barely sharper than the Tamron 24-70 zoom and that lens is the softest lens i got by a mile, and i did not even include it in the start OP because it is so much softer than the others lenses.







Those who were to utilize and shoot a fast f1.8D lens @f2.8 it would be a disaster regarding sharpness even on a 36mp sensor, this is a huge difference and the f1.8D is on par with my Tamron Zoom that is waaaaay off in sharpness for a D800 sensor, and it get even more magnified on a D850.

Wrong in so many ways it's hard to know where to start.

Left Sigma 50 Art

right Nikkor 50/f1.2

The Nikkor 50f1.2 is quite close to the Sigma 50 Art regarding sharpness @f2, the Nikkor is a clear keeper for those who got one.

Center of frame both @f2



Left side @f2





@ f5 center



@f5 left

Left Milvus 85/f1.4

Right Nikon 85/f1.4D



Zeiss is sharper, more contrast, better colors and no CA. The Nikon 85/f1.4D is not up for the job on D850 even at f8 its not sharp to the sides, and color and contrast is not good on Nikon.

Center @f1.4



right side @f1.4







Center @f8





Left@f8, heavy green and purple CA on Nikon

I'm still waiting on my D850.

But my friend has been shooting the hell out of his and I've seen his comparison shots with the D810.

The resolution advantage is plainly visible even when using the 50mm AF-D. Resolution using the mid-priced 85mm 1.8 AF-S is stunning.

Tons of lenses, old and new, will have a resolution advantage on a D850 body.

Rob

An unfair comparison, 70-200E FL +1.4TC and 300/f2.8

The zoom is only 280mm, (not even that i suspect ), i assume the nearly 20year old 300/f2.8 AF-s D is sharpest even if they were framed equally, and that by a good distance to the 70-200/f2.8E FL. Old BIG primes have not evolved much regarding sharpness, they vere perfect 20 years ago. As commonly known TC`s have some degrading effects as seen on the 70-200E and 200/f2, but they work ok i think.

I did not test my 400/f2.8, but i know its perfect, much better than the 200-400 zoom and way better than anything with TC`s.









Lower right corner:

left 200f2+1.4TC

Right 300/f2.8 (20 year old lens)







200-400/f4 (2004-2010 year)

300/f2.8 AF-s

The 300 Prime is equaly sharp at 3oomm as the 2-400/f4 is at 400mm !!

Both crops is 200%, but @f2.8 on the 3oomm







2x Tc on 70-200 vs 200-400/f4

MrHollywood said:

I'm still waiting on my D850.

But my friend has been shooting the hell out of his and I've seen his comparison shots with the D810.

The resolution advantage is plainly visible even when using the 50mm AF-D. Resolution using the mid-priced 85mm 1.8 AF-S is stunning.

Tons of lenses, old and new, will have a resolution advantage on a D850 body.

Rob

Click to expand...

Obvious all lenses will get better, but its sharply diminishing returns. I still stand by my conclusion that regarding sharpness it is better to buy new lenses on the 36mp than to use money on a D850. Regarding sharpness i would rather buy a Nikon 70-200E FL on a D800e than to uppgrade to an D850+keeping the old Tamron 70-200VC, that is obvious.

The D850 have other more interesting features, but sharpness over 36mp is not on the top of my list. By all means the D850 is today the best FF/FX camera on the market, so i bought it to do mainly landscape, birds and night photos.

Sum up, the D850 may have only a handfull (or non) lenses that can resolve its sensor from 200mm and downwards. To put a Nikon 50/f1.4 or any 24-70 zoom will not resolve the sensor by far. But all BIG primes i have used have been perfect, even the 20year old 300/f2.8 here, big primes basically resolve the sensor.

I stopped reading after the word "Important".....

north w said:

Important, this writing will contain hyperbole's and big words that is only related to a image that is about 6,2ft wide. If one only see whole pictures at computers or small prints as A2 or A1 this writing is not very relevant and the reader should skip this thread unless pixel peeping is of interest.

Click to expand...

That was good advice.

Reilly Diefenbach said:

Wrong in so many ways it's hard to know where to start.

Click to expand...

Who can argue "Reilly Diefenbach" ? not messurements of DXO it seems !

Point in case, i have proven measurement from DXO that your statement regarding sharpness of the 50/f1.8D is wrong, please stay off commenting when you dont have anything informal to share on the thread.

north w said:

MrHollywood said:

I'm still waiting on my D850.

But my friend has been shooting the hell out of his and I've seen his comparison shots with the D810.

The resolution advantage is plainly visible even when using the 50mm AF-D. Resolution using the mid-priced 85mm 1.8 AF-S is stunning.

Tons of lenses, old and new, will have a resolution advantage on a D850 body.

Rob

Click to expand...

Obvious all lenses will get better, but its sharply diminishing returns. I still stand by my conclusion that regarding sharpness it is better to buy new lenses on the 36mp than to use money on a D850. Regarding sharpness i would rather buy a Nikon 70-200E FL on a D800e than to uppgrade to an D850+keeping the old Tamron 70-200VC, that is obvious.

The D850 have other more interesting features, but sharpness over 36mp is not on the top of my list. By all means the D850 is today the best FF/FX camera on the market, so i bought it to do mainly landscape, birds and night photos.

Sum up, the D850 may have only a handfull (or non) lenses that can resolve its sensor from 200mm and downwards. To put a Nikon 50/f1.4 or any 24-70 zoom will not resolve the sensor by far. But all BIG primes i have used have been perfect, even the 20year old 300/f2.8 here, big primes basically resolve the sensor.

Click to expand...

Then why can I see the difference with a old 50mm 1,8 AF-D?

My 50mm and 85m G lenses are even sharper.

Beyond that the D850 offers more than a bump in resolution. The D800 was so-so for AF and the D810 was a nice improvement. The D850 AF is truly excellent. And that's but one example.

The D850 is a far more substantial bump in overall features and quality than we've seen since the D3 and D300 were first introduced.

Rob

Now, that's a bit rude, don't you think?

First things first - before you start typing some reply ready to blast me to the moon

I pretty much agree with most of what you've typed, but I have a large bone to pick with you in your attempt to prove that you're right and Reilly is wrong. Hopefully you'll take this post as a learning experience and not as an attack.

It's important, I think, to put out some disclaimers: I shoot D800E bodies, and I'm known for evaluating lenses subjectively for a very high standard of image quality, mostly because I've shot large format in the past and my reference standard is a combination of what an 8x10" large format chrome looks like on a lightbox and what a large gallery quality print looks like from large format film. However, I print *much* smaller than you do - most of my work sits at 13x19 to A2, and occasionally I might go to 20x30 inches. So what I print, to you, is a tiny little postcard! However, I've taken the time to do one thing that cuts a lot of the BS out of forum discussions - the blind print test, and I've found, for 24, 35 and 85mm lenses (I've only done this with those focal lengths - proper blind print testing is one royal pain in the backside, but now I have a good idea of what real world results are), that the majority of viewers CAN tell the difference between pretty good and great lenses, even when stopped down to F/9, at A2 (or 17x22") size, viewed at normal viewing distances, assuming the subject matter has fine enough structures to show detail. So even though we have different print sizes, we both print, and I think we both have arrived at similar conclusions. Obviously if I printed at your sizes, I think the magnitude of difference between lenses would be more noticeable in print tests.

So we're on the same page there, I think you'd agree.

Where I have seriously strong disagreement with you is in your use of DXO measurements to prove your case versus Reilly. Frankly, DXO's perceptual megapixel measurement is one of the worst things to come along, and people are using it to prove arguments and make decisions that may or may not be relevant to their use cases. We have to understand that no lens designer uses DXO perceptual megapixel ratings as a goal when designing lenses - I don't believe it's an accepted practice at all in the lens design community. What DXO is trying to do makes sense on the surface - in a fast food, ATM, instant gratification culture, people gravitate to simple answers, simple ratings, single numbers. Unfortunately lens performance CAN NOT be summed up in a single score, and never will be. If one takes the time to learn at least some of the basics of lens evaluation, it will become clear why this is so. Anyway, the DXO perceptual megapixel score is a weighted score across the apertures, done from their properietary target analysis method (the DXO blur analyzer) done at closer distances on a 2D target. Not particularly relevant to landscape photography done at long distances, and given they don't do a great job of disclosing their weighting, it actually doesn't tell us much about how a lens is doing at a specific aperture, particularly with regards to how it's resolution is at various structures. So the gold standard for measuring lenses is of course optical bench MTF, and unfortunately those are expensive, and very few sites provide such info. We are lucky that Roger Cicala and the guys at OLAF, particularly optical design student Brandon Dube, have given us insight into this world through the lens rentals dot com blog, with lots of MTF discussion, and it's valuable reading and education for sure.

If we think about things in terms of MTF, and we have to start with that I'm unable to find any measured MTF from the 50/1.8D Reilly speaks of, is that we have of course coarse structures (commonly thought of as contrast, in MTF terms the 10lp/mm traces), moderate, and fine structures. Fine structures these days are generally thought of as in the 40 to 50 lp/mm range, although there is some thought that in the higher megapixel camers, we could report higher. We have diffraction of course, which on a D800E starts to creep in (roughly) above or around F/6.3 or so, and while I haven't shot a D850 yet, probably F/5.6 or so on that body. But it's not a hard wall, I don't think. We have the nyquist frequency of the sensor, which is of course a hard wall to consider. But the important part - and the part that might explain why both you AND Reilly can be "right" is when we get to the much harder question of "What level of contrast at XX frequency is good enough to be considered tack sharp?", or "How much contrast at each structure type is enough". And there, of course, you, I, and Reilly might have different answers entirely. SMPTE states that a rule of thumb (that some, including myself is a bit excessive) is to have 80% contrast at half nyquist frequency. Well, good luck with that in most consumer lenses, on a D8xx class body across the frame. So that definition might leave us with only the Otus and similar at a non diffraction limited aperture, and certainly leave us skipping the 100$ 50mm lenses from the late 80's for sure. But the key here is that we, each one of us, likely have our *own* standards, based upon our experience, and most definitely because of our *output*, of what is "sharp enough". As stated before, my thoughts are more along the lines of yours, but I see where Reilly is coming from *if* I "adjust" for his (different than my own) standards for sharpness. We could also begin to discuss how *much* actual resolution there *is* (available) in any given scene/scenario, but this post is running long, and I'm not sure I'll make it under the character limit, so that's another discussion for another day.

I do think your reply to him was a bit unnecessary in tone, and I would caution you upon relying on DXO perceptual megapixel measurements to argue your case. There is so much more to lens evaluation than a single number.

I'll let the rest of you continue the conversation, just wanted to insert my two cents here. Thanks for the screen shots, although your results don't surprise me at all.

(For your reading pleasure, something that you and I will never afford, but the MTF of the panavision primo 70 cine lenses is quite a bit better than what we can mount on DSLR bodies: https://www.smpte.org/sites/default/files/24-1100-TS1-2-C&P01-Caldwell.pdf)

-m

I learned something just this evening while fine tuning and testing a 45mm & 85mm Tamron that arrived this afternoon. I was certain that my 85 was decentered or whatever because the lower left target in particular was way off. I was about to write in for another B&H return, but went to reread Lensrentals article on lens testing.


Roger suggested that if you get a bad edge reading shoot the targets again. But this time in portrait mode. So I did that and yes the bad target moved. But it moved to the wrong place. So I went to making sure everything was in line (like he said) to less than a half an inch and re-shot and walla the lens is actually centered pretty darn good. It was my camera alignment that was off a few inches. Seems I tend towards reading the instructions only after I mess up.

So now I'm thinking that not only do I now need to consider when reading about someone else's lens sharpness test if it was done with a lens or lenses fine tuned for that distance. I now also need to consider if their edge sharpness was done with a close to perfectly squared setup.

Wonder if you tried the flipping to portrait hold trick before giving up on that lens?

PS: So far I haven't had any lens come close to not showing a large improvement between shooting with the D850 compared to the D5. But I don't have any lenses I'd consider bad. Even the 150-600mm G2 shows a great improvement when shot with the D850.