User ID

Gave up on micro tuning AF ages ago. Whenever I need really exact AF I just use live view. LV is magic for MF also.

johnlutz Loc: florida

just left a gymnastic event and had the same problem.
went to my D750 and it worked perfectly. Must have a problem understanding the focus app on this 850?

therwol Loc: USA

That's fine for stationary subjects using AF-S. The D850 is lousy for following moving subjects in the Live View mode, even something as simple as a child in a playground. The contrast AF hunts for focus as distance changes, going in and out of focus. This is very well demonstrated in this YouTube video about using the D850 for shooting video where you're basically in Live View and looking at the subject on the camera's screen. See what happens when the narrator walks around. To me, this is the Achilles heel of this otherwise very fine camera. For what you pay, the video/Live View performance should match the performance you get when shooting with the Phase Detect autofocus that the camera uses with the mirror down. The only Nikon DSLR that uses Phase Detect in the Live View mode is the D780. If the D850 were ever to be upgraded, I'm sure this would have been incorporated into the new camera, however given Nikon's emphasis on mirrorless now, I doubt if the D850 will be replaced by another DSLR.

https://www.youtube.com/watch?v=s9AzriajNEM&ab_channel=ScottWydenKivowitz

therwol Loc: USA

So what do you do when the AF calibration value is a lot different at both ends of a zoom lens, like your 70-200? Do you really stop and change the value when you zoom in or zoom out? Do you put the value in the camera half way between the two values?

User ID

Thaz why I avoid the D850 or anything like it in the circumstances such as you describe. I use full time LV cameras for super accurate exact focus in a hurry.

I’d avoid using a view camera in such circumstances. It’s too slow. My 5D-SR replaced my view cameras. It’s slightly less slow, waaaay more portable, and hugely cheaper to operate.

When the a7R mkV ships I can replace the 5D-SR with an a7 mkIV. But as of right now it’s just too pricey.

TriX Loc: Raleigh, NC

Canons (and I assume Nikons) allow 2 settings for a zoom and you can pick the points - the short and long end being typical, but you can pick any two FLs you like. Everything in between is of course a compromise in that it assumes the error is linear, which may not be true for all lenses. Some lenses like Tamrons allow multiple calibration points using a “tap in console”. The same is true for distance - the correction is slightly different at 5’ and 20’ (I actually tested this and I’ll add a link to the page). But here’s the point: even with the correction error for varying FLs and distances, the corrected AF is ALWAYS better than the uncorrected AF.

Now the counter argument is that if you send the lens AND the body to the manufacturer’s service center, they may have more calibration points available. My personal experience was that with Canon, a lens that needed a pretty substantial correction (+13) was returned by the service center with the comment that it was within specs (which wasn’t good enough for me). Maybe Nikon is more precise… that’s actually what started me calibrating lenses, but take into account that I consult as the QA manager for an aerospace company that makes flight qualified components, so we live and breathe calibrated instruments and precision measurements, so I’m anal on the subject of things being 100% correct, not just good enough.

live view isn’t the answer. If you have time for that, you might as well magnify 10x and manually focus.

therwol Loc: USA

As far as I can tell, Nikon bodies allow for one fine tune adjustment per lens. I've heard about the Tamron dock, and I'm going to assume that the fine tune adjustments are applied in the lens. But I don't own any Tamron lenses.

Yes, Live View on Nikon DSLR bodies leaves a lot to be desired. The Live View on the D780 with phase detection focus may be an exception. I don't have any personal experience with it. Any mirrorless camera that has phase detection is going to be much better, and that includes some of the Nikon mirrorless bodies. I'm not at a place in my life where I want to go mirrorless and replace a bunch of lenses I have that won't work properly.

So here is where I am. I have a D850 and some interest in video, simply because I have two grand kids growing up, and video is much better to capture an "experience" than stills. The kids can grow up and watch themselves do things and hear themselves talk, laugh etc. Using this camera is out. Not good enough. The 8 year old HD camcorder I have is questionable. It has very poor low light performance. I'm thinking of buying into the low end of the higher end camcorders and call it a day, maybe spending a couple of grand. I like the ergonomics of a camcorder for video, the way it is held, the motorized zoom etc.

Grahame Loc: Fiji

I would suggest that is a far too simple procedure he demonstrated 'of a principal' that doesn't address three basic fundamentals if you are looking for the best accuracy along with confidence of results;

a) Although he has placed the camera focus box over an area of the sloping target he would not know for certain if the camera is focusing to the front, centre or rear of that box area (or possibly outside). Those three points having three different distances to the sensor.

b) Every control system has a 'deadband', which is a range within which the system makes no changes/adjustments so as to alleviate 'hunting'. If there was no deadband the system would never stop attempting to correct itself.

For this reason it would be practice to undertake a test shot with the lens initially defocused to the front and then another with the lens initially defocused to the rear. Difference in result would be assessed.

c) A number of tests, say three, are generally undertaken and assessed for consistency of results before any conclusion is jumped to and adjustments made.

TriX Loc: Raleigh, NC

Which is exactly what the FoCal system does. (Defocuses, focuses multiple times with different corrections and plots a curve of acuity vs correction): https://www.uglyhedgehog.com/t-508300-1.html

Terkat

Good morning TriX,

I found your reference to the FoCal system interesting. I did some investigating / research and think I just might give that system a "try" (no pun intended "Tri" - well maybe a very little pun).

All the best to you and yours in the upcoming year,

Terry