Before going further I am going to flag up a D5/500 relative weakness for bird photography.
With a D7200/810 you can use any AF point you want including multi points when focus tracking.
With a D500/5
with a lens like the 500 AF only the central 15 cross type AF points work as distinct from illuminating in the viewfinder
limiting tracking a moving subject to a form of centre weighted screen coverage.
The limitations are on pages 98-101 of the D500 Manual.
While not directly related to the topic I suggest any "serious" bird photographer should consider this D5/500 limitation as part of an upgrade decision.
You keep saying you must have a good target. I do not think you are right. It helps but there are ways that don't require the ultimate.
I agree in the field "the ultimate" is not always possible.
On the other hand when testing it helps to use a subject where AF is likely to perform well.
instead just wanted to evaluate my AF performance on the D500. I normally shoot birds. Therefore I need a typical bird as a target. I cant find one that will sit still for 2 hours so I get a cardboard box with some good lettering/images and a steel ruler. Most birds are unhappy if you put a ruler next to them. Set up the D500/500mmF4 on a reasonably steady tripod. Take 10 shots focussing on a good contrast area(focussing to infinity in between each shot) and then look at the field in focus on the steel rule at 45 degrees. With spot on focus there should be the same distance on the ruler in focus in front of the focal plane to what is behind. What I am interested in is how much does the center of the focal range move forward or back between each shot.
With a "perfect target in perfect test conditions" (not real world) there should be negligible variation.
A problem with a 45 degree rule coming from infinity is AF locks on the first area of contrast it can read, which may be a little further away than the intended mark on the rule.
What the variation ratio might be could depend on how steady the camera/lens is locked on a tripod.
It could also vary with the camera body. The D500 has more bur smaller AF points than the D800/810. My experience is the D500 can accurately lock on smaller areas of detail than the D810.
The D500 AF could be capable of locking on further out from the intended line than the 810, resulting in more inconsistency with this type of test. A variation like this does not make either camera AF good or bad. It shows they work different with some subjects. Which is best for a particular type of subject is down to experience and to the photographer.
….But when I am shooting birds, I don't get the chance to say 'your not a good target, can you move round a bit'. I am stuck with what I am shooting. I expect the body to give me a good focus at least 90% of the time or more. My experience with Nikon DSLRs over the past 4 years is that their AF is good enough to focus on very low contrast brown birds in bushes providing branches dont get in the way. A cardboard box is much better.
I agree AF is constantly getting better, particularly with low contrast subjects.
Current AF is significantly more capable than with my F100 in 1999.
Leonard, when you shoot birds, what % of shots do you expect in focus – for a static bird at 5-15m ?
How long is a piece of string?
Static birds large in the frame in good light are easy - around 99%.
Large birds which fly slowly, like a stork, often around 90%.
Small acrobatic birds like swallows mobbing a cat - very low - but worth trying.
Swallows feeding young while in flight on a predictable flight path during the half second they hang in the air to put food in a chicks mouth maybe 80%, ignoring perfect timing, framing and is the bird facing the right way
Digressing again
in the UK it is relatively easy to "change" the odds for around £200 a day.
In July you can get gannets diving for fish from a small boat having paid the boatman to throw fish into the sea nearby. The gannets soon learn which boats bring an easy meal and when.
At one location trout are kept in a pool surrounded by reeds - and a photography platform. The "guide" with a walkie-talkie advises which direction an Osprey is coming from and when, making it relatively easy at 10 fps to capture an otherwise once in 5 years chance of an Osprey taking a fish from water.
There are UK "set-ups" for kingfisher taking fish from water, or a Sparrow Hawk feeding on a goldfinch - with the goldfinch nailed to a photogenic branch!
I appreciate this is not "wildlife photography" as you do it, and it requires relatively little of an AF system. Even so it does take place.
This way of testing however, lets you see just how repeatable your combo is. What I saw today was somewhere in the region of 10% of shots outside of the focal plane I was aiming for (ie the front to back of the field in view was forward of the average centre point for that fine tune setting). That was based on around 100 shots. Last time I did this with the 800E my recollection was that the 800E was better (none out of that range?).
What you are trying to measure is the often very important difference between technical AF efficiency and your "real world" use in less than perfect shooting conditions.
The bottom line however is that the combo must focus quickly and accurately when the next Baird's sandpiper turns up as it is likely to be a lifetime shot. I cannot think – oh, the target isn't on the list that Nikon recommend for this body!!!! (and I don't think a Baird's sandpiper is on that list). Knowing the capability of the body on poor targets is important to me and this test method helps me understand how often I can expect perfect performance.
What you perhaps have not mentioned is knowledge at the basic level of what a Baird's Sandpiper looks like, or at the higher level of where you might see one, in addition to knowing exactly how your equipment works the way you shoot based on regular usage.
Given your experience of these DSLR's – and you have been around the DR forum as long as I have – how do you think the repeatability of getting the target focal plane in the field in focus is - with varying target contrast?
I still have not found an AF system that can detect the bright yellow but lacking any contrast or leaf separation in a globe flower (an uncommon form of UK buttercup). Whether this summer the D500 can do it remains to be seen.
But, yes - it would be good to know definitively whether temperature makes a significant enough difference.
