Nikon D850: the Focusing Distance Issues Stem from a Sensor that is not Parallel to the Lens Mount (Nikon quality control issue)
A few days ago in Nikon D850: Apparent Build Tolerance Error in Flange Focal Distance vs Optical Viewfinder, Causes Major Focus Error with Manual Focus Lenses and Focus Confirmation, I reported on a disturbing problem with the loaner Nikon D850 that I was using for testing.
The issue was/is that all of my Zeiss lenses were focusing well beyond infinity on the D850, most needing to be racked to their mechanical focusing limit. Yet none behaved that way on my Nikon D810 when making comparisons, and none were ever off like that on my D810 in the past 3 years.
Reader Stefan P had no joy with his D850, finding a similar issue:
Thank you for your recent information on the Nikon D850.
I tested my Otus 55mm on it and the results were the same as reported. As a consequence I returned the D850 today with the printout addressing these issues. I have 5 of the lenses listed in your blog and am not prepared to keep the camera under the circumstances.
As an aside, several of the camera staff read the blog but there was only one that appeared to understand it.
I did rework my report to make it easier for the layman (which apparently includes camera store staff!). Hopefully it is easier to understand now.
Tomorrow, I should receive a Nikon D850 of my own (well, I hope it is tomorrow). I am hoping that it does not have similar quality control issues.
Root cause: sensor-to-mount misalignment
I can now say with high confidence now that the root cause of this is not just a sensor to lens mount flange distance issue, but a far worse problem: the sensor is not parallel to the lens mount. That’s the worst possible scenario: a camera problem that degrades every image taken with every lens used. Even stopped down as far as f/5.6, latent unsharpness can be seen in some cases—this is not a minor matter.
With this kind of issue, certain focus positions will show strong image degradation, but others may show just puzzlingly low performance in some areas of the frame, because the focus may in essence balance out the badness across the frame.
I went back and looked at the images I had taken, and a pattern was seen: images focused in the center taken on the D850 were moderately to badly blurred on the right side of the frame. Images taken on the D810 performed as expected—no symmetry issues.
Diagnosing sensor to lens mount misalignment
The trick to diagnosing such issues is to shoot two or more known-good lenses on a known-good camera and the suspect camera.
To test for sensor misalignment, choose a subject that is at far away, ideally 5000 times the focal length (e.g., 400 meters for an 85mm lens), so that there is no depth of field or alignment issue (close range raises issues of being at 90° to the subject as well as the subject varying in distance as well as differing behavior at close vs far range). The subject should have fine detail at a uniform distance across the frame, particularly at left and right.
Here is my protocol, done for each camera:
- Mount lens on camera, enter 100% Live View and focus/shoot at 3 positions wide open: far left, center, far right.
- Repeat for each camera + lens combination.
- Load all images into layers in Photoshop; examine them for symmetry.
When I did this with the Sigma 85mm f/1.4 DG HSM Art and the Sigma 135mm f/1.8 DG HSM Art, the Nikon D810 delivers symmetric sharpness across the frame, with no meaningful deviation in sharpness across the frame for the 3 focus positions.
When I did this with the Nikon D850, the left/right focus positions resulted in moderate to major blur on the opposite side of the frame.
While one in theory could have two lenses that are skewed inversely to the way the camera is skewed (thus hiding the issue) that’s just for giggles and my Nikon D810 is a proven performer for 3 years. And the idea that two brand-new lenses would be off just perfectly to offset a camera skew is just not realistic.
I deem this Nikon D850 issue a quality control problem. Variable quality control explains why some readers see no issue, and some like Stefan P above have similar findings: erratic build quality due to sloppy tolerances. Cameras off that much should be adjusted prior to leaving the factory—basic quality control for a $3400 camera.
The good news for anyone with this issue: Nikon service ought to be able to adjust the sensor to lens mount alignment. That’s exactly what I did some years back with my Nikon D800E; it was out of alignment just like this D850 (same symptoms), but when it came back from Nikon the blur problems were gone.
Toggle to compare. Similar degradation is seen on the D850 on the right side of the image when focused on the left side, but the D810 shows no meaningful difference.