D800 AF sensors - here's my result, and how I tested

Started Apr 17, 2012 | Discussions thread
Flat view
Julian Vrieslander Senior Member • Posts: 1,057
D800 AF sensors - here's my result, and how I tested


D800 s/n 3003xxx gives poor results from the outer AF sensors, with the leftmost sensor faring the worst. The center cross sensor, and contrast detection AF (in Live View) appear to work OK. This seems to be similar to a pattern reported by others. I am posting this note to share my results, and to describe my testing method, which should be easy to duplicate.

If you want to skip to the sample images, here is the link:



I created some AF test targets similar to one that has been used at Nikon service centers. In another post, I described the target and provided a link to a PDF file, so you can print your own copies:


I mounted the D800 on a tripod in front of a large glass window. I used the reflection in the window to align the camera's optical axis perpendicular to the window. Then I taped focus targets to the window at positions which appeared under the AF sensors.

I used a 24-70mm f/2.8G, zoomed to 50mm. Camera was 2.5m from the plane of the targets.

AF mode: AF-S
AF area mode: single point
Release mode: Mup
Exposure: 800 ISO, f/2.8, 1/80 sec.
Lighting: mixed incandescent and CFL, not as uniform as I would like

Shutter was tripped via remote release. Before each shot, the lens was defocused at infinity. I made three shots for each of the tested AF sensors, and there was no significant shot-to-shot variation.

I tried a different test procedure with the camera at the same position, but shooting only the middle target, and rotating the camera on the tripod to place the tested AF sensor over that target. This produced the same results, with good focus from the center AF sensor, large errors from the leftmost sensor, and smaller errors from the rightmost sensor.

Using AF-C mode, or setting the initial defocus to a near position, instead of infinity, did not change the results. I also did some checks to ensure that the AF sensors were not being influenced by features surrounding the targets, behind the window, etc. Because of time limitations, I did not test all AF sensors - only the three mentioned.

I repeated the entire test series with a 50mm f/1.4G, and found again that the leftmost AF sensor failed to produce an acceptable focus. But with this lens, the rightmost sensor gave acceptable results, performing better than it did with the 24-70. The center sensor focused correctly.


Shots made with the center AF sensor always produced the sharpest images. The AF errors of the leftmost sensor were significant, repeatable, and present on both of the lenses tested. Accuracy of the rightmost sensor was somewhere between those extremes.

On all the shots, even those made with the center sensor, I could see some smaller variations in sharpness across the frame. Some of these effects could be due to slightly decentered lens elements, imperfect parallelism between image sensor and lens mount, etc. The D800 is going to be a nightmare for fussy pixel-peepers. They will need to take great care when attempting to isolate the sources of optical flaws.

I don't know what is causing the errors from the outer AF sensors. Since other people are reporting satisfactory results from all sensors, there may be sample-to-sample variations. My first guess is that, for some units, there is a flaw within the AF sensor, or that it is not aligned correctly. I sent my sample images to Nikon, and I'm waiting for a response.

Nikon D800
If you believe there are incorrect tags, please send us this post using our feedback form.
Flat view
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow