Absolutely stunning performance

[kenw]

This lens got mixed reviews on release and more than one reviewer did note their lens was "decentered" to some degree meaning one or more corners/edges were softer than the other side. This is of course depending on the degree a manufacturing defect and not something you want hear.

I recently purchased a 14-30 and so wanted to immediately test it for any decentering. One must have reasonable expectations here as a zoom will almost always show some, especially in the extreme corners, at some part of its zoom range.

The test I used is similar to that described on Jim Kasson's blog here:

Lens screening testing - the last word

Note: the best way to navigate this set of pages is to use the "Articles" section in the right-hand panel. It has links to all the pages and their titles. For the last few years, I have followed Internet discussions related to lens screening. By screening, I mean the process that a photographer...

blog.kasson.com

Basically it is to put a Siemens Star target at a good distance from the camera and shoot the target at the center and the corners. In Kasson's case he makes it totally idiot proof by putting the camera so far away that the center and corners will all be within the DoF when you recompose to put the target in the corners. This requires a very large target and often extreme distances from the target. My test is similar except to keep the shooting distance reasonable for indoor work I calculate how much to change the target distance when moving from the center to the corners. I do not refocus on the target after the distance change. Thus I'm still testing for a flat field but don't have to be at outdoor distances to do this. I just need to do a little math in a spreadsheet to get the distances correct for flat field testing.

I shot at 14mm, 20mm and 30mm. I shot at F4 (wide open) and F11 (a typical aperture for landscape work). I shot both 75% of the way to the corners and at the extreme corners.

Test was on a Z7 shooting RAW and processed in Adobe Lightroom. No vignetting correction was applied. In order to view the results similar to how I would shoot a landscape the "Detail" tab settings for "Sharpening" were set to my typical starting point which is Amount 50, Radius 0.8, Detail 80.

I'm pretty blown away by the results:




F4 Top row 14mm, middle 20mm, bottom 30mm. Center and 75% towards each corner.




F11 Top row 14mm, middle 20mm, bottom 30mm. Center and 75% towards each corner.




F4 Top row 14mm, middle 20mm, bottom 30mm. Center and absolute extreme corners.




F11 Top row 14mm, middle 20mm, bottom 30mm. Center and absolute extreme corners.

These are 100% crops from a 45MP sensor. Note in the F4 shots the lens is resolving so well we are actually still seeing some Moire in the extreme corners. At F11 the entire field including the extreme corners is essentially diffraction limited.

We can see the heavy vignetting at 14/4 which is not a surprise and reported elsewhere.

For the extreme corners the stars are not perfectly centered because it is extremely difficult to position precisely. For the shots 75% to the corner I repositioned the crops slightly to center each star but for the extreme corners I wanted to keep the crops on the exact corner.

If you aren't familiar with how these kinds of tests should look check out the link above where Jim has a whole bunch of sample lens tests.

In my book this is pretty phenomenal performance. The field appears to be nearly perfectly flat and this testing is far enough from the target (200 times the focal length) to likely represent infinity performance closely.

Perhaps I got extremely lucky with my "copy" of the lens. Whatever the case my expectations were exceeded by a wide margin. This is amazing good and consistent performance across the entire field. I'm very happy with the lens!

--
Ken W
See profile for equipment list

 

[James Stirling]

Thanks for posting your results Ken, I have been delighted with my 14-30mm , pretty small and light, gives great results and the bonus of taking filters without an adapter faff. I had not seen your posts in this forum, I am sure your knowledgable and reasoned posts will be most welcome here as they are in the m43 forum.

Your results mirror my experiences with the lens . I love UWA lenses and have used the 14-24mm since it came out. I was just saying in another thread that though I have some small uses for a faster UWA the upcoming 14-24mm Z will need to be very special indeed as the 14-30 on my Z7 is a lovely combination especially for wandering the hills and mountains

--
Jim Stirling:
It is not reason which is the guide of life, but custom. David Hume

 

[syberman7]

In Kasson's case he makes it totally idiot proof by putting the camera so far away that the center and corners will all be within the DoF when you recompose to put the target in the corners. This requires a very large target and often extreme distances from the target.

It's not that hard when testing 14mm at f/4. I copied Jim's instructions to the letter and tested using the 410mm target at 9m from the camera. The only exception was that I used two fluoro soft-boxes for lighting.

My results aren't as sharp as yours, but they are not performed quite the same way. Maybe Jim could explain. Below are the 14S at 14mm f/4, and the Sigma 14mm f/1.8 at f/4 from the same position. The Sigma is clearly superior to the Nikon, which isn't surprising considering that it's a large and heavy prime with a bulbous front element. For Astro work I always use the Sigma. I think the Sigma results validate the testing method as well.





For the Nikon Z the performance in the corners picks up quite a lot at 18mm:


Nikon Z 14-30 @ 18mm f/4 - 10m from chart

I'm quite happy with my specimen the 14-30 as the charts look reasonably consistent across the four corners.There's no obvious dud chart rendering. But, like I said, at 14mm there are sharper options if you can bother.

Thanks for your post.

 

[third son]

Thanks for the summary. This lens is next on my list.

 

[Rich Rosen]

Had it since July. Placed in a local county photo contest with one of the first photos I took with it.







--

Rich Rosen

https://www.flickr.com/photos/richardrosenphotography

 

[kenw]

Thanks for posting your results Ken, I have been delighted with my 14-30mm , pretty small and light, gives great results and the bonus of taking filters without an adapter faff. I had not seen your posts in this forum, I am sure your knowledgable and reasoned posts will be most welcome here as they are in the m43 forum.

Nice to see you over here too, James. Yes, it would appear after many years I've finally succumbed full frame-itis. Nikon appears to have ticked the boxes I was looking for that Canon and Sony have so far not really excelled at.

The forum over here seems vaguely familiar to the m43 forum. The frequent posts about the impending demise of Nikon makes me feel right at home!

Your results mirror my experiences with the lens . I love UWA lenses and have used the 14-24mm since it came out. I was just saying in another thread that though I have some small uses for a faster UWA the upcoming 14-24mm Z will need to be very special indeed as the 14-30 on my Z7 is a lovely combination especially for wandering the hills and mountains

Yes it seems like there would be a very small set of lenses that bring much more to the table!

 

[kenw]

In Kasson's case he makes it totally idiot proof by putting the camera so far away that the center and corners will all be within the DoF when you recompose to put the target in the corners. This requires a very large target and often extreme distances from the target.

It's not that hard when testing 14mm at f/4. I copied Jim's instructions to the letter and tested using the 410mm target at 9m from the camera. The only exception was that I used two fluoro soft-boxes for lighting.

Yes, indeed with this short a focal length following Jim's method exactly still leads to quite reasonable working distances. I really only used my method here just so my results were a bit more comparable with other tests I was doing on other lenses at longer focal lengths.

My results aren't as sharp as yours, but they are not performed quite the same way. Maybe Jim could explain.

I'd say your results at 14/4 appear only a little different from mine in the extreme corners.

I have a theory as to why that is. In Jim's test the effective position of the test target is following a curved field rather than a flat field as you move it around the frame. His helpful charts give the recommended minimum distance for which the effectively curved field fits within the DoF for a given aperture and pixel pitch. But DoF is by no means an all or nothing proposition, being at the edge of it still reduces sharpness. With Jim's charts we expect this reduction to be small and to not be severe enough to mask the "decentering" for which the test is designed to detect. But it still can certainly be potentially visible at these 100% views. So assuming a lens has a truly flat field there is the potential my method will show slightly sharper corners than Jim's.

Now then the Sigma looks awesome in your test, as it should being a wonderful prime quite a bit stopped down by F/4. But it might be getting an additional boost in your test because according to a few reviewers the Sigma does exhibit some field curvature. Since in Jim's test the "virtual" test target we create by panning the camera around is actually curved and not flat it is potentially following the field curvature of the Sigma better than in the Nikon which may be more flat than the Sigma.

And one step further it is quite possible that the Nikon isn't actually flat but actually slightly curved in the opposite direction of the Sigma! Usually field curvature curves "inward" - that is it looks concave from the direction of the camera. But there have been a few UWA zooms that have actually show "outward" field curvature. If that were to be true than my more flat virtual target would make the Nikon corners look better than your test with slightly inward curving test field even if our lenses were performing identically.

This is relatively easy to check for. As Jim suggests you can run a second test in which you refocus when shooting the corners (only refocus once on one corner and then shoot all four corners at the same focus). If the corners sharpen up when you do this then you know the lens field is a different shape than test target field.

Or of course at least as likely the different copies of our lenses perform slightly differently!
Thanks so much for sharing your test images. Nice to see reasonable confirmation of my results and neat to see the impressive performance of the Sigma prime. Naturally when I'm saying this lenses greatly exceeded my expectations that's down to my expectations of it being a slowish UWA zoom!

 

[syberman7]

This is relatively easy to check for. As Jim suggests you can run a second test in which you refocus when shooting the corners (only refocus once on one corner and then shoot all four corners at the same focus). If the corners sharpen up when you do this then you know the lens field is a different shape than test target field.

Thanks for your thoughts Ken. I have tried the refocus on the corners technique. I came to the conclusion that it was barely worthwhile. It made some charts looks better and some worse. I think its the short throw focus by wire system used in the S lenses. When in manual focus magnified view with peaking engaged, the slightest turn of the focus ring moves the peaking up and down the converging lines of the star in a very jerky fashion. It's very difficult to finesse fine focus. Do you find the same?

And another thing; if a lens has decentered elements the field curvature will be tilted. I think it gets very difficult to interpret what's going on if you refocus for the corners.

Just as an example here's my 85S with centre focus and 'lock', versus refocus for every chart:




Centre focus only




Refocus each chart position

What do you think?

 

[kenw]

This is relatively easy to check for. As Jim suggests you can run a second test in which you refocus when shooting the corners (only refocus once on one corner and then shoot all four corners at the same focus). If the corners sharpen up when you do this then you know the lens field is a different shape than test target field.

Thanks for your thoughts Ken. I have tried the refocus on the corners technique. I came to the conclusion that it was barely worthwhile. It made some charts looks better and some worse. I think its the short throw focus by wire system used in the S lenses. When in manual focus magnified view with peaking engaged, the slightest turn of the focus ring moves the peaking up and down the converging lines of the star in a very jerky fashion. It's very difficult to finesse fine focus. Do you find the same?

Yes that's my experience as well. To be clear I actually don't have a lot of experience with the Z system yet but it is my experience with other focus by wire systems as well. And a similar problem exists with many mechanical AF lenses that often have very short throw and with consumer lenses are even a little sloppy with some backlash.

What I've found with long experience in the m43 system and my brief experience now with the Z system is that CDAF on a high contrast test target like this is exceedingly accurate. There are no "offsets" or "backfocus" and "frontfocus" issues with CDAF unlike with the PDAF systems on DLSRs we've been trained not to trust. Thus I think Jim's recommendation in his write up to use manual focus doesn't actually apply here. I think forcing CDAF by selecting Pinpoint AF is likely to give better results on the Z system than doing MF given the vagaries of the focus by wire that you describe.

For my test shots I found the most confidence in focus was by putting on magnified live view, using Pinpoint AF and then hitting the AF-ON button while in that magnified live view and seeing it snap crisply into focus with moire everywhere.

Now CDAF can certainly get thrown for a loop in real world scenes just like any AF method so I still expect to be checking things with magnified live view and peaking in my real world shooting. For these Seimens Stars though I suspect Pinpoint AF is going to be easier and possibly even more consistent than MF with zoomed peaking.

And another thing; if a lens has decentered elements the field curvature will be tilted. I think it gets very difficult to interpret what's going on if you refocus for the corners.

Absolutely agree. I'd only refocus on a corner if the corners already looked equal in the center focus only test. That is to say I already know there is no decentering or tilting effects but now I'm just curious what the story is with field curvature. And again here I wouldn't refocus on every corner, I would just refocus on one corner and then shoot the other three at that same focus setting.

If there was decentering or tilting visible then of course it is sort of impossible to choose which corner to use for the refocus. And refocusing every corner individually isn't really particularly informative at this point either.

Just as an example here's my 85S with centre focus and 'lock', versus refocus for every chart:


Centre focus only


Refocus each chart position

What do you think?

Agree with these there is very little to distinguish them at all. Clearly little point to refocusing.

I have in the m43 world had a few cases where I had a lens that investigating field curvature was of utility and refocusing on one corner bringing all four corners into focus showed quite clearly that field curvature was present. The Olympus 12/2 prime has a bit of it, enough that it is still visible at F/5.6 on a 16MP sensor. With that lens I learned that if I cared about detail across the entire field that if I set the focus point about 1/2 to 2/3 of the way to corner this balanced out the defocus from field curvature between the center and the corners (if I recall this idea was from Anders W who used to post a lot in the m43 forum) . Interestingly when I was just looking up whether the Sigma 14/1.8 has field curvature there is a similar recommendation from Nasim at PhotographyLife for that lens. He suggests using live view to check both the center and corner as you adjust focus manually to balance out the field curvature in the 14/1.8. The Nikon Z's unique Split Screen liveview would make this pretty efficient.

--
Ken W
See profile for equipment list

 

[syberman7]

For my test shots I found the most confidence in focus was by putting on magnified live view, using Pinpoint AF and then hitting the AF-ON button while in that magnified live view and seeing it snap crisply into focus with moire everywhere.

Great suggestion. I'll give that a try next time.

 

[James Stirling]

Thanks for posting your results Ken, I have been delighted with my 14-30mm , pretty small and light, gives great results and the bonus of taking filters without an adapter faff. I had not seen your posts in this forum, I am sure your knowledgable and reasoned posts will be most welcome here as they are in the m43 forum.

Nice to see you over here too, James. Yes, it would appear after many years I've finally succumbed full frame-itis. Nikon appears to have ticked the boxes I was looking for that Canon and Sony have so far not really excelled at.

The forum over here seems vaguely familiar to the m43 forum. The frequent posts about the impending demise of Nikon makes me feel right at home!

Yep, it seems that like newspapers bad news is good news for generating interest. Our choices have put us on the potential double disaster road

Your results mirror my experiences with the lens . I love UWA lenses and have used the 14-24mm since it came out. I was just saying in another thread that though I have some small uses for a faster UWA the upcoming 14-24mm Z will need to be very special indeed as the 14-30 on my Z7 is a lovely combination especially for wandering the hills and mountains

Yes it seems like there would be a very small set of lenses that bring much more to the table!

I have enjoyed the tiny Samyang 35mm F/2.8 on the Sony mount. A surprisingly decent performer given its size and price. Alas though it will mount at AF on the TZE-01 the aperture does not play nice. I see Nikon have a couple of compact primes on the roadmap hope they come fairly soon

 

[Rolf_52]

I think your experience mirrors the experience of most users. The average rating at Adorama based on 42 reviews is 4.9 and the reviews on BH-photo and Amazon are close to that.

Personally I hesitated because of mixed reviews but finally ordered and after my usual testing found no signs of decentering. In fact, the lens was jaw dropping good given the zoom range, size and price.

My Z 24-70 f4 is very good at 24 mm but my 14-30 is slightly sharper and at 35 mm my 14-30 is close to S 35 mm. I also remember a thread where a user had difficulty seeing an advantage of the new Z 24 mm f1.8 over this little wonder.

 

[kenw]

In fact, the lens was jaw dropping good given the zoom range, size and price.

Well put. Its the context that makes it amazingly good.

My Z 24-70 f4 is very good at 24 mm but my 14-30 is slightly sharper.

I see this as well. I was pretty impressed with the 24-70/F4 at 24 until I saw what the 14-30 was doing. They overlap quite nicely with the 14-30 being just a bit better in the corners from 24-30 than the 24-70/F4 and by 30-35 the 24-70 is looking amazing. But this is all at extreme pixel peeping levels. The 24-70 is already excellent on its own from 24-30.

 

[contrelamontre]

(...)

My Z 24-70 f4 is very good at 24 mm but my 14-30 is slightly sharper.

I see this as well.(...)

+ 2. Particularly at the 24mm range (which is very important for my use cases).

 

[Rolf_52]

In fact, the lens was jaw dropping good given the zoom range, size and price.

Well put. Its the context that makes it amazingly good.

My Z 24-70 f4 is very good at 24 mm but my 14-30 is slightly sharper.

I see this as well. I was pretty impressed with the 24-70/F4 at 24 until I saw what the 14-30 was doing. They overlap quite nicely with the 14-30 being just a bit better in the corners from 24-30 than the 24-70/F4 and by 30-35 the 24-70 is looking amazing. But this is all at extreme pixel peeping levels. The 24-70 is already excellent on its own from 24-30.

Thanks for the comment. I agree you only notice the difference at extreme pixel peeping. They are both superb lenses.

 

[TristanW]

Dude, did you try 14mm RAW without correction?

Thanks!

 

[kenw]

Dude, did you try 14mm RAW without correction?

Thanks!

The test shots above are without vignette correction but they still have distortion correction on. I also presume they have lateral CA correction with the built in profile but I’ve not double checked that.

The 14-30 has lots of distortion with corrections removed, you can see the amount in this review:

https://photographylife.com/reviews/nikon-z-14-30mm-f4-s/3

I’ve been shooting EVF only cameras for so long now I’m just used to lenses where the designers properly ignore the constraints imposed by the presence of an OVF on distortion. I expect any EVF only camera system to potentially have a fair bit of distortion and lateral CA present and a built in profile automatically applied by RAW converters.

I always think of a profile as one more lens element in the design. I don’t use or test my lenses after removing the rear element of the lens so I don’t usually test or use them without their profiles either

Vignetting though I can’t make my mind up on entirely for these kinds of tests. If there is a lot of vignetting that adds noise in the corners when you correct for it which isn’t obvious in these base ISO well exposed high contrast tests. So I do want to know about the vignetting. One the other hand it is harder to visually compare sharpness and contrast in the corners if I don’t have vignetting corrected. I sort of feel like I should probably have both...

And some people point out you can make a lens a bit wider by turning off distortion correction if your subject can tolerate it. But that’s not a normal use so I don’t bother testing it myself. I just look at other reviews that check uncorrected distortion and make a mental note that it might be an option. So far in a decade of shooting lenses that often have a lot of uncorrected distortion I’ve never actually needed to do that though!

EDIT: Clarification on CA correction. It appears that of course Horshack has already posted a deep dive into the applied corrections in the Z system here:

https://www.dpreview.com/forums/post/62193330

And his post shows that lateral CA correction is part of the built in lens profile. That said I see my default option on import into LR is to also check the "Remove Chromatic Aberration" box. In some of the extreme corners checking and unchecking this box does actually show improvement in CA with it checked which suggests there is some small remaining lateral CA that the built in profile doesn't correct perfectly.

 

[Mach Schnell]

I have to agree on the absolutely stunning performance. I bought the lens a couple months back. I did some tests for decentering (not as elaborate as yours) and found nothing amiss. Then I headed off to shoot. Here is the first real-world image I captured. The lens is a definite keeper. This is 4,128 pixels across, or exactly 1/2 the resolution of my Z7.







--
Steve

http://www.pbase.com/fotofanatik/image/73553598.jpg

 

[Clayton1985]

Dude, did you try 14mm RAW without correction?

Thanks!

I doubt the OP tested the lens without corrections or without removing the lens cap since neither is very useful for actually taking pictures with the lens.

 

[Mackiesback]

I have to agree on the absolutely stunning performance. I bought the lens a couple months back. I did some tests for decentering (not as elaborate as yours) and found nothing amiss. Then I headed off to shoot. Here is the first real-world image I captured. The lens is a definite keeper. This is 4,128 pixels across, or exactly 1/2 the resolution of my Z7.

I saw your words and this picture, which looks like Zion and my first thought was, does this guy live in St George?



Turns out.......



Well done sir.