Quality nikon wide glass

[NicholasRab]

Looking for some peoples thoughts,

I have very little experience with Nikon, having shot Canon and Hasselblad for, well, forever. Not sure if I have ever shot a frame with a Nikon slr.

I've been planning for quite a while to move to a new medium format tech cam system this spring, however if the rumors are true and the D800 does turn out to be a 36mp camera, I may try to work with that.

95% of my shooting is on a tripod, stopped down, mirror lock up, etc. With that consideration I have some questions and concerns about the nikon wide angles... will they be up to it?

The 24 and 45 t/s would be core lenses. Do they hold up well at 24mp? I find the canon 24t/s quite good, is it comparable? What about wider? It seems as if the 14-24 gets good reviews. Is this good in a "good for a ultrawide zoom" or good in absolute terms?

I'm a little confused here, as I was ready to shell out a bunch of money for a more traditional architecture setup, but it seems like (if it happens of course) a D800 with 14-24, 24t/s, 45t/s, and 70-200 would do what I need for far less and with a far easier workflow (live view, etc).

 

[Vivek]

No one can answer your question w.r.t to the not yet released D900.

The answer to the query on 14-24 is, it is one of the best ultrawide/wide lens (prime or zoom) in the market.

 

[kuau]

Nicholas,
I can tell you from first hand experience, the nikon 24mm t/s lens will not perform well or the 45 t/s on the yet to be released D800 36mp camera, I used both lenses on the D3x and there performance were sub par at best and I used the same shooting technique as you describe.
All I can say is when Nikon releases there 36mp camera, Zeiss mf lenses is what I would be looking at.
If you want t/s movement, hi res, then MF tech view is the only way to go
steven

 

[NicholasRab]

Thanks for the thoughts.

Vivek - I tried (possibly unsuccessfully!) to make it clear that I am not really asking any questions about the upcoming camera, more to the performance of wides on the existing 24mp platforms. If they are out-resolving the sensor in the corner at 24mp that would point to their performance at higher resolutions.

Kuau - Not what I wanted to hear, but I am not necessarily surprised. Thanks for your experience.

 

[kuau]

Nicholas,
sorry to bring the bad news, though vivek is correct the 14-24 is excellent I had one when I had my D3x

 

[Vivek]

Nicholas, Sorry I overlooked the information you were looking for.

But the second post did catch my attention. How would you know when a lens outresolves a sensor in a camera?

 

[kit laughlin]

The Nikkor 12-24 is one of the absolute best lenses I have ever used, and as sharp as the Leica–Zeiss 15 I used to use on a 1Ds and 5D.

I have not used it on a 24Mp camera (tho' how I would have loved to on the Sony A850), there are absolutely no surprises with it (apart from some simple barrel distortion at the 14mm end, and only on objects a metre or so from the front of the lens.

From this I infer that it will only make better images on a 24Mp sensor.

I'm intrigued by that part of your post too: [when the lens is]

out-resolving the sensor

How would one know that?

In any case, the 14–24 is very good in an absolute sense, better than Nikon's primes in the same focal lengths. I remember when the MTF charts were released, some of the lens pundits on FM were even saying that, somehow, Nikon had cheated physics with that performance.

As an aside, I switched over from Canon with Leica/Zeiss glass on the day the D3 and the 14–24/2.8 was announced, and had one of the first of these kits in the country.

Last point: kuau is 100% spot on, if you want the best quality. The Nikkor 24 PC-E disappoints, IMHO, until ƒ11, and has none of the Canon's flexibility in movements. Only you can decide whether the 14–24 can give you what you need in this regard. I can say that I have images coming out in a prestigious European house/architecture magazine mid-2012, using a bunch of images I took with the D3s and D700, and the 14–24 (processed in Aperture, too, for Jono!). Cheers, Kit

 

[NicholasRab]

Hi Guys,

As far as "outresolving the sensor", maybe I should have said the opposite. Its obvious when a lens is not able to provide enough resolution for a particular sensor. The corners are soft, etc.

I also realize that there are people doing fantastic work with far inferior equipment than what we are speaking of. In truth I could probably continue using my 5d2 along with the 17 and 24 t/s lenses. While there are times that I could use more resolution for printing large, those needs are fairly infrequent.

As for the Nikon 24mm t/s, that's a bit of a shame. I can't say that I have ever shot the canon version I have under f/11, but it has been a very dependable lens.

 

[routlaw]

Just ran across this thread and feel the need to respond. Understand i am going to be the contrarian on just about everything stated here. Hopefully non of this will sound snarky, or denigrating, its not meant to be.

First, lets talk about 36 mp sensors pixel pitch size on the Dxxx or what ever. I suspect you could stuff all of the lenses made world wide in your t-shirt pocket that can truly, effectively resolve a pixel size this small. And even they existed, would probably only be made by Rodenstock and Schneider, certainly not Canon or Nikon. Worse still would be the limited DOF due to diffraction loss at a much larger aperture, and man you better have your shooting technique down to science as this will not be horse shoes and dominos when it comes focusing in the field. At one time on both rodenstock and schneiders website they had white papers on the resolving power of their digital/digitar lenses, non of them resolved beyond 5 microns and make no mistake packing 36 million pixels into a 35mm full frame sensor will have dramatically smaller pixels, probably something along the lines of 3-3.5 microns.

Nikons PC-e lenses: I have all 3 and they are all great lenses with a few caveats mechanically which I will come to. I am still using them on my D3 but have demoed them also on a D3x and found the only limiting optical factor to any of them on this 6 micron pixel pitch sensor was the OPERATOR, not the lens. Rather than go into any great detail about this I would suggest you subscribe to Lloyd Chambers advanced photography reviews where he extensively and I do mean extensively reviews all of this equipment together and in short has nothing but praise for it but with some limitations not the least of which is diffraction loss beyond certain apertures due to the 6 micron sensor of the D3x. I don't have any connection with LC, but will guarantee it will be the best $40 bucks you spend all year. It is worth noting that Lloyd emphatically claims the only way you can obtain absolute focus with these lenses on this type of camera is with live focus, and while I have succeeded in accurately focusing these lenses otherwise will have to admit he makes a very valid point. I have tested my 45/85mm PC-e lenses against the Zeiss 50 macro, a very nice lens indeed, and found the Zeiss to have better corner to corner resolution, but the difference was rather subtle than jaw dropping as some would claim. It is worth noting that I had each of these lenses dialed in with my cameras fine focus feature, shooting a very detailed test target. Zeiss does have a different look and mechanically they are a joy to use but you can also get fabulous results with the PC-E lenses so long as they are used with meticulous technique.

Nikon 12-24 Lens: Tried 3 of them on 3 different bodies, D2x, and 2 D3's and rejected all three lenses due to focusing issues being so inconsistent, not only at a given focal length but the focus fine tune would change dramatically at different focal lengths. Effectively what this meant was saving some half dozen different FFT (focus fine tune) settings in the menus for different focal lengths. And if this was not bad enough those settings would change depending upon whether you were close focusing vs infinity or half way across the room. All 3 of the 12-24 lenses exhibited this trait and they all had widely different serial numbers. That said when you got it right the images were nice and amazingly even at the widest focal length barrel distortion was almost non existent. No small feat.

To the original poster it would well behoove you to read other reviews on these 3 PC-E lenses, including Lloyd Chambers www.diglloyd.com reviews of them. Celebrated landscape photographer Jack Dykinga uses a D3x, D3s, and all 3 of these great lenses for his work over the last few years. Given he spent his entire career with an Arca Swiss doing 4x5's I seriously doubt he would be using this system if it were not providing him with equal or better performance. One of the photo magazines ran an article on him a couple of years ago you might find this an interesting and informative read.

Finally why not rent a D3x and all 3 of these lenses from lens rental dot com and find out for yourself, but only having read through Chambers method for getting the most out of the system. This will be chump change compared to what the system would cost.

Hope this helps.

Rob

 

[kuau]

Rob,
I agree with everything you have to say about the 36mp yet to be released nikon
yet in regards to the d3x and nikons 3 pce lenses, I used to own this setup and yes I am a Lloyd subscriber, and actually spoke with him on the phone a few times about the d3x and nikons pce lenses, I found only the 85mm pce lens to be a top performer.
I sent all 3 of my pce lenses to nikon to have to t/s put on the same axis and had them check my 24 and 45 for any optical issues and they tested at spec.
I was very meticulous in technique, tripod, mlu, cable release, live view focus etc.
my experience with the 24 pce at f8 and f11 with no shift iq was good, but as soon as I added some shift corners suffered and I had the same experience with the 45 pce.
I spoke to Lloyd about my findings and he had the same issues with both the 24 ans 45 pce.
These lenses all came out when the D3 was introduced and having a 12mp FX sensor with large pixels, all 3 lenses performed very well, I can verify this because I used a d700 with all 3 lenses and had no issues. D3x different story.
I admit I do to much 100% pixel peeping in the corners and probably 95% of the time when doing prints from my d3x, the final results were very good.

When 36mp FF cameras finally do ship, it will be very interesting to see if Nikon or Sony have newer lenses that can do them justice.


steven

 

[kit laughlin]

Rob, I was debating with myself as to whether to re-post on this; I did the conversion on axis orientation myself, and like kuau, I am a subscriber to Lloyd's site, and used the recommended techniques for the sharpest images, and found the same as Steven did with the 24 (I was using rise and fall more often, as well as tilt for infinite DOF, and assessing this via Live View).

Perhaps I exaggerated a tiny bit when I said that this lens 'disappoints under ƒ11'; it would have been more accurate to say 'underwhelmed'—definitely very good, but did not "wow" like some many images from the 14–24/2,8 did, wide open and up to ~ƒ5.6. Again, this is just my experience and I may have had an exceptional copy of that lens.

Your advice to the original poster is totally sound, if I may say, and it's something I must try myself (tho' this is much harder to do here than where you live). Cheers and thanks, kl

 

[D&A]

I really can't contribute much more than has been said already here in this thread, but my fairly lengthy experience with the 24 T/S (and the brief time spent with the 45 and 85) when shifted, was also underwhelming when shot on a D3x. As Steven and others have pointed out, used on the D3, images were more than acceptable. Unfortunately, shortcomings in edge and corner details were quite evident when used on the D3x, especially when any siginificant degree of shift was applied. This also applied to some other highly regarded lenses which previously preformed well with 12MP bodies, but not with the 24MP D3x.

The samples I've owned of the 14-24 f2.8 were all uniformly excellent and were up to the task on the D3x in my opinion. It's a spectacular wide angle lens at all focal lengths with high resolving power and low distortion. By far the best wide angle lens/zoom I've ever shot with on 35mm. The Zeiss 21mm arguably equals or comes close to it.

As Steven has pointed out, acceptability in image quality with these lenses/body combinations, is greatly dependent on output use. Web? Print media? (1-2 page spreads), or sizeable large format prints?

I couldn't agree more with Rob's suggestion to rent out the bodies and/or lens combinations you plan on purchasing, to see if they are up to your specific requirements...and also subscribe to Digilloyd's site. Sometimes it's often subjective and what is acceptable to one, isn't to another.

I too wonder about current lens performance in general, if/when we see a 36MP DSLR from Nikon.

Dave (D&A)

 

[D&A]

To Add to my Post Directly Above...

The reason I mentioned subjectivity as to what's acceptable and what's not, I'll use Nikon's 16-35mm f4 VR lens as an example. It's overwhelmingly liked by most who own or use it, yet it is one of my least favorite wide angle zoom lenses (putting it mildly). The level of distortion between 16 and 20mm is completely unacceptable for my particular use (and one of the highest I've ever encountered in a wide angle lens/zoom. I am not refering to normal wide angle stretch on side/edges of frame, thats to be expected. Yes, many say simply correct for its high level of distortion in post processing...but its not that simple. Ever try to shoot group "Grip and Grin" in moderately close interior quarters with it? Yes, the horendous facial/body distortion can be corrected at the expense of having the all the straight lines/walls and identifiable items in the room completely distort. Can't have it both ways...unless maybe work hours on a single image, and thats a big "maybe"! When there is well over 300 or more images like this, the 16-35 VR lens for that and other types of applications is unworkable. Yet for those that use it for landscapes or subjects with straight lines that distort, then for them, the lens may be the one to pick. Just an example of how subjective it can be. Therefore if at all possible, rent or try beforehand.

Dave (D&A)

 

[routlaw]

Just a few more comments to add to the conversation and another possible suggestion to the OP on this thread. For me the D3x was somewhat of a love hate situation when I demoed the camera.

Much to my surprise general IQ looked better than my D3 regardless of the lens assuming I had all my ducks in a row and as previously stated that was not always the case. Effectively I found the extreme fussiness with which one had to operate this camera to get the most out of it just wore me down. And it should be noted I have a reputation for being very picky meticulous person. Lloyd Chambers alluded to this also in his reviews comparing the D3x to the D3. In other words unless you had everything 100% dialed in you might as well have used the D3 because IQ (at least in terms of resolution detail) got tossed out the window. Using my Ebony 4x5 by comparison is far more pleasant experience. I mention this because with the D800/36 mp camera this situation will become far more extreme than the D3x was and then there is the pathetic USB 2 connection to computer which will be very slow compared to a FW800 connection that the MFDB tend to use. Transfer of files from the D3x in the studio was very slow compared to the D3. Why Nikon has insisted on this type of connection is beyond me.

Now all this brings up the D4, which I have placed an advanced NPS order for and feel this will be the right set of compromises or attributes for me. Higher resolution than the D3, but most likely no where near as finicky for dialing in IQ compared to the D3x, faster transfer speeds in the studio due to some new connections and apparently even higher dynamic range than either of the other two cameras. The 7.3 micron sensor will not place nearly as high of a demand on the glass in front of it, plus diffraction limits will not kick in nearly as quick compared to the 5.5 ± sensor of the D3x, yet it will be much better at eliminating moire for my studio shooting with certain products.

Taken a step further if I need more resolution, there is always pano stitching frames with the D4. A 3-4 frame stitch would provide more resolution than any 36 mp camera could hope for and do so without putting so much demand on the glass while allowing for far better DOF for these types of images, ie landscape, architecture & interiors etc.

Most comments about this new camera seemingly have it categorized as a sports/action/wildlife type of camera only and at the risk of being proverbially contentious and contrarian I could not disagree more. Yes it will be great for that type of photography but it will and can suit needs far beyond that for which I have just described. It would be unfortunate to see photographers eliminate this new camera from consideration based upon such narrow criteria as sports/wildlife only capabilities. I'm certainly not.

Ok thats my two cents worth.

Rob

 

[NicholasRab]

Appreciate all of the thoughtful discussion here. A rental and comparison between my 5d2 and a d3x on t/s lenses would probably be very worthwhile. I find the Nikon interface a bit bewildering regardless (not having really used it), and some time spent with it would really let me know how functional it would be for me. Note I find the Canon interface ridiculous as well, It's just a ridiculousness I am familiar with. Guess I am a little old school with my dislike of modal controls.

I also did a little more research into the Nikon t/s lenses, and it looks like they are a bit less flexible than the EOS ones, in regards to axis manipulation. I still find the construction of the latest gen canon t/s lenses pretty amazing.

I think that, as usual, a wait and see approach makes sense here. 95% of my commercial output is either printed very small or used on the web. The hangup is that last 5%. When I need to run a huge print, the 5d2 is okay, but not great (and my, how my standards have changed since film). Of course, when I do need to run that print, its for the high paying client that is crucial to keep happy, and they want it cropped a little bit, etc. In addition in the last couple years I went from a Hasselblad H3D-39 to a 5D2, and its hard, seeing the resolution difference on paper. My images are better from the Canon in absolute terms, and that's what truly matters, but I'd love to have that resolution back.

Again, thanks for the excellent discussion.

 

[danielmoore]

Given the 4.7µm pixel pitch on the current APS-C sensors, a 36 megapixel camera would be simply using a larger surface area to produce that file. I suspect the better glass will be a good match provided one adhere to the diffraction limits, which without getting into it that whole quagmire, should be close to f/8, a very workable combination.
For my own work I'm very much looking forward to a higher res camera for cropability alone. The Nikon 14-24 is so good, in many cases I'm happy to level the camera and crop an image to maintain perspective control, having twice the linear resolution is going to be dreamy. For Architectural work, Helicon Focus is a godsend, achieving better overall results than I have using T/S, which still only manage planes of focus. Helicon Remote takes a lot the work out of your hands as well, making it much less tedious and in the end possibly even faster than setting up a T/S required scene.

 

[kuau]

Daniel,
You are probably right about the 4.7µm pixel pitch on the 36mp sensor,
Though from what I have been seeing and reading about the NEX-7 anything under a 28mm lens is an issue, so I assume it will be even a bigger issue on a FF sensor.
Helicon Remote looks very interesting, and maybe even faster then using tilt, yet for architecture and landscape, I still think shift is a great feature to have...

Nicholas, Yep, I think the new V2 of the canon TS/E lens is better then Nikons. though when I check out DigiLloyds tests using the the newer 24 TS/E on a 5D MKII, i don't think the results are that great.

Steven

 

[NicholasRab]

Steven, I could be mistaken, but I think that a lot of the problems with the NEX-7 and wide angles has to do with non-retrofocal lenses and angle of light upon sensors. As practically all SLR lenses are retrofocal I think that the wide angle problems of the mirrorless crowd are not related.

Daniel - I have not used Helicon in a few years, since I sold my MF gear. Sounds like it's been improved, I'll have to take a look.

 

[D&A]

>>>Steven, I could be mistaken, but I think that a lot of the problems with the NEX-7 and wide angles has to do with non-retrofocal lenses and angle of light upon sensors. As practically all SLR lenses are retrofocal I think that the wide angle problems of the mirrorless crowd are not related.<<<

This is my understanding too as to why ultra wide angle lenses on the Nex and some other mirrorless cameras pose a problem.

With the upcoming full frame D800 supposedly having 36MP...why shouldn't it pose even more of a problem with many lenses than the full frame D3x? (as was suggested it wouldn't). Thats the question I'm trying to figure out and why if its anything like dealing with the D3x... a lot of shooters won't be coming close to the potential of such a camera.

Dave (D&A)

 

[routlaw]

Daniel with all due respect I think your logic and math are off on these conclusions. First assuming all the rumors of a D800 with 36 mp are correct and we're not all whistling in the wind on this conversation it should be compared not to an APS sensor but a full 35mm frame sensor like the D3 series of cameras. The pixel size of the D3x is 5.9 microns @ 24 ± MP, so the D800 would be an increase of a full 50% percent of pixels in the exact same area. According to my calculator the D800 would have a 3.96 micron pixel dimension, far short of the 4.7 you mention. For field work I have relied on the rough equation of 1.5 x the micron dimension of the pixel to figure diffraction limits and this has served me well for years. Many other pros have done the same. In any case this brings the diffraction limits of F6 ± a 1/3rd of an f-stop for the D800. Using this method in the studio while demoing the D3x photographing resolute targets with the best of macro lenses sure enough diffraction started kicking in right past F8, noting F9 would be 1.5 x the 5.9 microns sensor size.

When I owned the D2x F8 was about as far as that camera would go, F11 and the images started to become semi worthless IMO.

Perhaps a couple of Nikon & Zeiss macro lenses might be able to resolve a 3.96 micron sensor but I would not put any money on the rest of the lineup for doing so.

Below are a couple of links if any one is interested in studying the laws of physics on diffraction, pixel size and lenses.

http://www.lonestardigital.com/aperture_diffraction_limits.htm

http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm

Hopefully all will find this helpful.

Regards

Rob