5th D800 from Amazon still has left AF issue!

Started Aug 23, 2012 | Discussions thread
D800
Regular MemberPosts: 199
Like?
Re: SEVEN in a row?
In reply to TOF guy, Aug 25, 2012

TOF guy wrote:

D800 wrote:

CFynn wrote:

D800 wrote:

In addition to these 5 D800s from Amazon, 2 more D800s from Bestbuy also had left AF issue.

Uhh - are you sure you are testing properly? Even if 50% of D800s had the left AF problem, the chances of getting seven cameras in a row with the problem would be very remote. The number of cameras with the problem is probably more like 10% to 20%, so the chances of getting seven in a row is like winning the lottery.

Simple answer, this left AF issue is prevalent . Some people got lucky with a 5th D800, some people tried several (up to 8 D800s) and still had no luck. Read here:

http://www.amazon.com/Amazon-delivery-Only-people-purchased/forum/Fx2MCFOKA9GJ1HG/Tx21XQKJ4UEQQ9H/35/ref=cm_cd_notf_message?_encoding=UTF8&asin=B0076AYNXM#Mx2UAUES43WMYSD

I think CFynn raises a very good point though. Let's assume 1 camera out of 2 is affected, (probably a "pessimistic" evaluation but let's assume the worse). Then the chance of "drawing" randomely 7 cameras which are all defective is about 1 out of 500,000. Nikon has stated that it plans a maximum throughput of 35,000 units per month. So we have of the order of 200,000 sold out there. There is less than a chance out of 2 for Nikon that one of its customers who tested 7 cameras ends up with all defective cameras.

Maybe the issue is more prevalent as you state. But maybe also the methodology you use to test the AF has a weak point, and therefore you get consistent results with a consistent method to test the cameras, but unfortunately the results only reflect the method used to test the cameras (same for the Amazon customer with 8 defective D800).

I've not very convinced myself of the above, but it's a fair point. Would you mind discussing how you proceeded to test the cameras (or point to a thread where you've done so? There are too many to keep track of !). In the "worse" case we may conclude that your approach if fine, all D800 tested have the defect, but you won't be any worse off. In the best case you'll be more adept at testing your next camera, whether it's a D800, a D600 or whatever.

-- hide signature --

Thierry

Test procedures are:
1. Manual procedure:

Camera is on Gitzo tripod and RRS ballhead, 2.5m from targets, camera's sensor parallel to the test targets, exposure delay 1s, F2.8, 1/200. Targets printed from Nikon 24-70mm service manual, rotate 90 degree and focus on horizontal stripes. If Leonard's targets used, distance is 27x focal length, focus on black rectangle.

a. Reset AF ring, take a reference image in LV with middle focus point, check for sharpness. Reset AF ring, take another image using PDAF with 0 AF fine tune, compare the sharpness of PDAF image with LVAF image. Change the AF fine tune value, reset focus ring and keep taking test shot until the sharpness of PDAF = LVAF, delete any image that does not have the same sharpness as LVAF. This is x which is optimal AF fine tune needed for middle focus point. At this point, I have 3 images: 1)LV with the optimal sharpness, 2)PDAF at 0 AF fine tune, a less sharpness than image 1, and 3) PDAF at x AF fine tune with has the same sharpness as image 1.

b. Everything is the same as in step a but now with left most AF point instead of middle. Reset AF ring, take a reference image in LV with left most point, check for sharpness. Reset AF ring, take another image using PDAF with x AF fine tune, compare the sharpness of PDAF image with LVAF image. Change the AF fine tune value, reset focus ring and keep taking test shot until the sharpness of PDAF = LVAF. This is y which is optimal AF fine tune needed for left most focus point. At this point, I have 3 more images: 4)LV with the optimal sharpness, 5)PDAF at x AF fine tune, a less sharpness than image 4, and 6) PDAF at y AF fine tune with has the same sharpness as the reference image 4.

c. Repeat same above procedure for the most right AF point. I have z value and 3 more images.

2. Semi-automatic procedure using Focal Pro with their recommended targets and distance. Confirm the optimal AF fine tune values needed for middle, left, right AF point closely match x,y,z respectively in procedure 1. See this link for more detail of this procedure: http://www.falklumo.com/lumolabs/articles/D800Focus/OuterSensors.html

3. Take the camera for a walk outside in daylight, with correct x AF fine tune value dialed in for that particular lens, shoot at everything I can find. Images taken with left PDAF point are OOF (back focus), images taken with LV left AF point and middle PDAF point are tack sharp.

If the difference of x and y is less than 5, I will consider that is acceptable and good copy. With 7 D800s that I own, the difference is always more than 10.

Let me know if you can find any flaws in my test procedures or anything that I did incorrectly.

Cheers!

Reply   Reply with quote   Complain
Post (hide subjects)Posted by
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow