Nikon D800 Contrast and Phase Detect Autofocus Testing

[box type=”info”]Update April 1, 2013: Testing after the second repair attempt by Nikon is now online here: https://photokaz.com/2013/04/nikon-d800-autofocus-repair-testing-the-sequel/[/box]

[box type=”info”]Update January 20, 2013: Initial testing showed changes to the focus performance and I thought the slight softness in some shots could be fixed with fine tuning.  After more testing, I could not get consistently sharp results from fine tuning. I had to set my fine tuning for the Nikkor 24-70 f/2.8 lens to +20 and that still didn’t result in good performance.  My AFS 50mm f/1.8 and AFD 80-200 f/2.8 lenses also had to have +20 of fine tuning.  I dropped off my camera and the three lenses with Nikon last week, told them the problems and said I wanted everything repaired.  I will once again post my results when I get the camera back.[/box]

[box type=”info”] Update August 23, 2012: I finally had a chance to do some critical tests of my fixed Nikon D800. I can say that Nikon did fix the problem (still to be tested), I have posted updated test charts here.[/box]

[box type=”info”] Update July 30, 2012: Picked up my D800 in Richmond today. I did some ‘off the cuff’ testing, hand held without a test chart. It seems to be fixed though I won’t know for sure until I set up a proper test procedure again. I won’t have time to do that until next week but will test again and post my results.[/box]

[box type=”info”] Update July 20, 2012: Nikon has confirmed that the camera arrived in Toronto. They provided a service order number and a link to check the status, unfortunately the page is ‘under construction’.  I hope that means it is just offline for a fix as I haven’t visited the page before.[/box]

[box type=”info”] Update July 18, 2012: I dropped off my D800 at the Nikon repair center in Vancouver (Richmond) along with my test results. They said that the high end bodies get shipped to Toronto, and I’ll be without the camera for two to three weeks. Not fun, but worth it if they can fix the issue. If it comes back the same or worse I won’t be too happy about it.[/box]

After reading several posts about the ‘left autofocus point’ problems of the Nikon D800, I decided to test my own camera to see if it needed repair.  Thom Hogan outlined a test method in his July 16, 2012 post and I decided to follow that method to test my camera, with a change.  Initially, I developed a test target as per Thom’s suggestion but it led to problems so I did a second round of testing with the classic Siemens star.

Some notes about the camera and test

• I have a camera purchased in Canada, it was very early in the release cycle so call it an ‘early run’ camera.
• I used two lenses for the test, the Nikkor 50mm f/1.8G and Nikkor 24-70 f/2.8.
• I have not made any changes to auto focus fine tune.
• I used a Really Right Stuff TVC-33 carbon fiber tripod and BH-55 ball head to hold the camera.
• All exposures shot in RAW, converted to jpg to display here but no sharpening, lens correction, etc. was added.  I only changed white balance and exposure to more easily compare things.  I also cropped to make comparison easier.  Everything managed with Adobe Lightroom 4.
• The lenses did not have any filters installed.
• Camera had distortion control turned off.
• Shot at the widest aperture (f/1.8 for the 50mm and f/2.8 for the 24-70) and ISO 100.  The 24-70mm was shot at 24mm.
• Camera was leveled using the built in leveling tool and a three-way hot shoe level.
• The camera (sensor plane) was about 6.5′ from the wall.

Test Results – 50mm f/1.8G

Top row is Live View manual focus, second row is Live View AF, and last two rows are viewfinder phase detect AF.  The 50mm does show less detail in the left AF sensor.  Note, bigger versions of these test images are below, this is just a summary.

Test Results – 24mm f/2.8G

The 24mm lens shows a much more severe left AF issue.

This is a tedious exercise to both shoot and post.  I did it twice as my first test chart didn’t yield useful results.  My D800 does suffer from the left autofocus problem when using phase detect autofocus with the two lenses I tested.  It will need to go to Nikon to get repaired.

Here is my test setup, in case anyone is curious.

Here are the full test patterns and should be in the same order shown above.