As indicated before, Mongo has waited 8 years to upgrade his camera body waiting for the quantum leap instead of the yearly incremental steps from his D200. So he bought a D800 about a month or so ago and about $3500 later. In many respects the camera disappoints but some are to scared to say so. More and more Mongo sees others on the net pointing out faults. They ring a familiar bell with Mongo.
Speaking to Sar recently, we both seem to have had at least one similar negative experience i.e. the focus seems to just "lock" at times and not refocus when you aim at something else or "freezes" on the object you last focused on and not refocus. Luckily, these were temporary "freezes" which thawed out moments later but annoying nonetheless and what Mongo considers is a design or factory fault !! Many others have reported the same experience on the net. Too many to be a coincidence.
Additionally, Mongo was unhappy with the sharpness/focus he was getting from time to time. There are countless owners who reported the same problem on the net world wide. The specific problem being that the left side AF sensors do not focus accurately. Generally, the middle and right have not been reported as being affected or as badly affected. Mongo did a quick test today to get it out of his head. Unfortunately ,that just put more steam in his head which whistled even more loudly from his ears.
Mongo's results show what he considers to be very bad AF focus from the left sensor(s) slowly improving across the middle sensor to the right sensor which is better again than the left (by far) and the middle (by a noticeable amount). Whilst the right sensor is the best of all, Mongo is in no way convinced that even it is functioning as accurately as it could or should be - it just happen to be the best of a bad lot.
3 test images appear below from far left to middle to right sensors. Mongo took several sets and they are all consistent with the set Mongo has attached here. Mongo thinks the difference is quite evident but becomes even more evident when you look at the first and then the last (skipping the middle image). Moreover, the original images show the differences even more dramatically on Mongo's screen at home.
The images have been standardised in that they have been taken and processed identically. They were taken in Jpeg and no sharpening (except "+1" which had been set in camera) or processing other than reduction for the web has been applied. In any event, Whatever has been done to one image,has been done identically to the others.
Test details: D800 on FX Jpeg fine, tripod, 70-200VR on aperture priority wide open at f2.8 set to 200mm, VR -off, single spot sensor used for AF, ISO 100, spot metering (but note that all exposures were identical in any event i.e 1/800th, f2.8, +0.7EV), distance to subject about 10 to 12 metres, all images were taken within seconds of each other.
Mongo used a 70-200 because of its accuracy and sharpness but set at 200mm. Apparently, this problem is far worse and noticeable when trying to use the AF sensors on a wide angle lens , say, wider than a 50mm.
You wonder how it could ever get through quality control and leave the factory in this state. Mongo has always said that if he wasn't so heavily invested in Nikon gear, he would be buying Canon. Mongo has had enough and it's going back to Nikon tomorrow.