44 years ago – Nikon Ai 20mm/3.5 Nikkor-UD Auto

I love old glass. Here is a special one for me. Introduced 1967 as the first ultra wideangle lens which could be normally operated on SLRs, the Ai 20mm/3.5 Nikkor-UD Auto. And now, 44 years later, it can be used on the latest generation of D-SLRs. That’s continuity.

Here is a short story.

Enjoy, Andy

Essay – Eiffeltower at Night

It is such an impressive monument. The Eiffel tower in Paris. Especially at night.
Earlier this year I had a chance to capture some night shots with the Nikon D700 and AFS 16-35mm/4G VR.

Here is the very short essay.


Update #2: Focus issues with pro-lenses

A quick update on my original post with the D7000 focus issues with the new high-end AFS lenses (24mm, 35mm, 85mm). Nikon support responded on friday afternoon, that they have opened a support case will look ito it early next week but would welcome more information.

After the initial surprise on friday, I did some tests at home yesterday and to be able to have some images to share with Nikon I did some shots at the Schönbrunn palace.

To cut a long story short. It is not a D7000 issue. It is a broader issue. Please don’t conclude from my experiences to a generic design or Q&A problem – there is just no evidence for this kind of speculation.

What is not speculative is the result of some test shots taken in the last 2 days with my Nikon gear and the action going forward.

Ok, lets start:

I am still suprised, that this “problem” of a significant focus inaccuracy with fast lenses shot wide open at long distance did not get my attention earlier in my life. Thinking about those years it is true, that I seemingly never shot under these conditions (fast lens, wide open, long distance). With fast I mean everything faster than f2. Ok, let me be more precise with my attestation: I’ve never shot with a high end fast lenses under this conditions. I did it with “cheaper” lenses like the AFS 35mm/1.8G and AFS 50mm/1.4G (or the AFD), but when I saw the medicore results it was easy to attribute it to the mainstream lens class –  as I was under the impression that there should be some difference between these classes.

Talking about wide open – free hand night photography is usually on short distance objects. Long distance night sceneries are usually taken from the tripod and stopped down. I did few images wide open and long distance. But with night photography these abberations did not get that clearly out, among all the other potential night issues.

Topic resolution. People used to view images at typical web resolution (900×600) won’t find any issues stemming from those high res sensors. If you don’t need more, stop investing your time in this article and move happily on. These artifacts are imho clearly visible above something like 2000  x 1300 pixel images (no hard numbers here). Every resulotion in between is dependent on the viewer.

Nikon positioned the new AFS 24mm/14.G and the AFS 35mm/1.4G explicitly as versatile tools coping with such broad shooting situations like portrait, still life , landscape and astro photography with “world class” low abberations for wide shots. With this kind of positioning, I didn’t feel completely out of place to use my lenses for shooting at distances of more than 100 ft (30m) wide open. At least “astro” sound awfully far away.

Some insights:

  • All 3 fast & new AFS lenses exhibit this focus inaccuracy and distortion with more cameras: I tested it on the D3, D3s, D3x, D700, D7000, D2x, D300 and D300s. I did not use the other consumer bodies, but there is no reason that it would work there either. Anyway, I wouldn’t want to use these lenses with the remaining bodies.
  • The AFS 24mm/1.4 is hit hardest by this problem at f1.4. At f2.8 most of the abberations are gone, at f5.6 it is on save grounds in this respect
  • The AFS 35mm/1.4 is following the 24mm closely
  • The AFS 85mm/1.4 is also affected at f1.4, but not as significant as the wide angle brethrens.
  • The AFS 200mm/2, one of the best Nikkor lenses does not AF properly with the D7000 at f2 (this is also true with the D3x)
  • The D7000 has AF issues with many more lenses (I did not test other camera/lens combinations): AFS 17-55mm/2.8, AFS 70-200mm/2.8 VR I, AFS 24-70/2.8, AFS 28-70/2.8 to name a few.
  • Based on its higher resolution sensor, the D3x is more affected that the other FX cameras (but they are affected as well)
  • I’ve also checked the AF fine tuning in close range with wide open lenses.

I did send Nikon some test photos to help their support team, but to me it looks like I have to ship a pretty big box to Nikon. I trust Nikon that they will figure out what is wrong here.

If any of the readers have images shot under similar conditions, I would love to hear from them and see those images (preferred NEFs), be it perfect in focus or not.  Many thanks.

If you are interested in the original NEF’s – please download them here. (take the image number as download index)


Some images:

Schönbrunn palace, all focus test images taken from a tripod, remote, mirror up, …
This was shot with the PC-D 85mm/2.8 – intentionally blurring the palace


Nikon D7000 (100% crops)


(D7K_3291) D7000 & AFS 24mm/1.4


(D2K_3294) D7000 & AFS 35mm/1.4

(D7K_3297) D7000 & AFS 85mm/1.4

(D7K_3300) D7000 & AFS 200mm/2


Nikon D3X (100% crops)

(D3X_1171) D3x & AFS 24mm

(D3X_1174) D3x & AFS 35mm

(D3X_1177) D3x & AFS 85mm

(D3X_1205) D3x & AFS 200mm/2


2 Images taken at a shorter distance.

As I said before: This problem is strictly a long distance/wide open issue between many camera bodies and lenses. All the mentioned bodies and lenses are able to deliver excellent images in different shooting styles.

D7000 & AFS 200mm/2 @f2, about 40 ft, handheld

100% crop


or from the zoo in Schönbrunn

100% crop



Finally – AF focus test with the spyder lenscal


D7000 & AFS 24mm


D7000 & AFS 35mm


D7000 & AFS 85mm


D3x & AFS 24mm


D3x & AFS 35mm

D3x & AFS 85mm


How much resolution do I need? 21 Nikon D-SLRs compared

Welcome to another round of comparison on this website. A while ago, I compared 21 Nikon D-SLR bodies out of my personal interest. How much resolution is necessary with the usual output devices and print sizes. Of course the level of post processing impacts results. As those requirements vary widely, I decided to upload all NEF files from the cameras, so that every interested reader can download the files from cameras he is interested in and apply his very personal post processing and scaling to the size he needs/wants. And THEN assess if he really needs the next resolution level.

I’ve tried very carefully to use the 21 bodies in similar ways in a stable setup to make the comparision as valid as possible. Covered bodies: D1, D1H, D1X, D2H, D2X, D3, D3S, D3X, D40X, D50, D60, D80, D90, D200, D300, D300S, D700, D3000, D3100, D5000, D7000 and as a special guest – P7000.

Further reading is in the article


The limits auf automatic longitudinal CA correction

Isn’t the current world of photo tools fantastic? Just one click in modern applications like Capture NX2 and artifacts like color aberrations are magically eliminated. At least, that is what we are made to believe. Those tools have some big disadvantages – they did not accompany you while you captured your photo, did not get an explanation what your intend was, did not recognize the objects in the scene. That’s why they are sometimes limited in their utility or even worse, counter productive. Let me share an example.

Two months ago, Europe experienced heavy snowfall, which impacted air travel in the northern part. I got stuck at Berlin Tegel airport. Finally, after 6 hrs delay and 2 re-bookings, we were allowed to walk to the Air-Berlin plane around midnight. I took a couple of images with the AF-S 35mm/1.4 mounted on the D700, while waiting for boarding over the staircase.

I took the image below and was really surprised when I developed the photo via CNX2 and wanted to save the final image on disk. This was not the image I recorded originally. It was significantly different. All the letters were framed in black borders. The red “Air Berlin” letters on white background, as well as the white letters on the red background on the engines.

Triggered by this abnormal behavior, I started to investigate for the root cause. Finally the culprit of this strange thing was found. Turning on longitudinal CA removal in CNX2 “created” this black frame around the letters. Turning it off, removed it. It is a repeatable thing. Weird.

I don’t have insight into Nikon’s software algorithms, but based on the behavior of CNX2 with my image, I would think that the way how this algorithm works is based on edge detection with certain color values and halo ranges. By selecting the “remove this” button, the software replaces the CA halo with darkened pixels, i.e black. To me it looked like the company colors of the Air Berlin plane was not in line with the automatic algorithms of Capture NX – hence the letters were framed black.

It is easy to reproduce with this image. If your are interested to try it for yourself – click here to download the original NEF file.


This the overall image

This is the image, when longitudinal CA removal is not used. (This is the way it should look like)

This is the image, when longitudinal CA removal is turned on. All letters are framed with a black border.

By the way, did I say I love the AFS 35mm/1.4 G ? I love it ….. (Taken from the staircase upon entering the plane)