This test doesn't seem practical. Using D153 for a subject that size? I could see that being useful for a bird against a clear blue sky, but in this test, you're using a shotgun to shoot an ant. In this situation 3D would've done what you wanted. Know, I know the D810 does... maybe that's wrong.
Here's an idea for a test to see if Dynamic Area does as described in the manual. I refer to a video or article that Steve had on checking the accuracy of the placement of an AF point, vs what's illuminated in the viewfinder.
Take a big piece of white paper with a little target test design in the middle.
In single AF point, determine the area or location of the actual AF point. Then go to D9 AF. Do the test again but as the main AF point goes off the test image and onto the white BG, one of the AF points surrounding it should still hold it in focus until you move away from the area that point can hold. The key here is (from the manual) - the camera will focus based on information from surrounding focus points if the subject briefly leaves the selected point. So if you stay off the target with the primary AF point, it's going to go back to it and then start hunting on the white BG.
In 3D mode. use the center AF point. As you pan away from the little test patter, the AF point will move and keep the test image in focus - it'll follow. English is my second language, but when I read the manual, that's what I gather. I don't know how or why the other bodies perform as people describe, but could it be that they aren't working properly as described?
Steve Perry wrote:
Yup, the camera rocks in single point and group - and that's how I use it most of the time as well. The Dynamic thing came up because of a project I was working on. Needed some sample shots and noticed it didn't work like in the old days.
I would love to know the outcome of all this. To be honest, I have never really given these modes much thought on how intricately they work as I have not had issue with them, but that was mainly on the D810 not the D500. As I said, I mainly use it in Single and Group thus the issue you talk about has not affected me
As for the AF-C issue, that - might - be easy Check A10 and make sure it's set to Off. I'll bet it's on AF-C.
Yes, it was set to AF-C only, but I wonder how as I have never touched that menu item! Thanks for the help, I also owe you a beer. However, I did put a donation to your wonderful site a month or so back.
I can confirm that same behavior on my D500 but I do not see it on my D750. Clearly there is an issue with Dynamic Area Mode on the D500. Have you tried contacting Nikon customer support to raise this issue?
la puffin wrote:
This test doesn't seem practical. Using D153 for a subject that size? I could see that being useful for a bird against a clear blue sky, but in this test, you're using a shotgun to shoot an ant.
Actually, that is a good point. It does seem a large AF area for such a small subject and I wonder if this has some significance. Maybe try 72 or even 25 and see if it happens for this same subject size? Maybe the AF areas are more responsive the close the Single point and lose effectiveness the further you get away from the Single point?
In this situation 3D would've done what you wanted. Know, I know the D810 does... maybe that's wrong.
Here's an idea for a test to see if Dynamic Area does as described in the manual. I refer to a video or article that Steve had on checking the accuracy of the placement of an AF point, vs what's illuminated in the viewfinder.
Take a big piece of white paper with a little target test design in the middle.
In single AF point, determine the area or location of the actual AF point. Then go to D9 AF. Do the test again but as the main AF point goes off the test image and onto the white BG, one of the AF points surrounding it should still hold it in focus until you move away from the area that point can hold.
In 3D mode. use the center AF point. As you pan away from the little test patter, the AF point will move and keep the test image in focus - it'll follow. English is my second language, but when I read the manual, that's what I gather. I don't know how or why the other bodies perform as people describe, but could it be that they aren't working properly as described?
la puffin wrote:
This test doesn't seem practical. Using D153 for a subject that size? I could see that being useful for a bird against a clear blue sky, but in this test, you're using a shotgun to shoot an ant. In this situation 3D would've done what you wanted. Know, I know the D810 does... maybe that's wrong.
Here's an idea for a test to see if Dynamic Area does as described in the manual. I refer to a video or article that Steve had on checking the accuracy of the placement of an AF point, vs what's illuminated in the viewfinder.
Take a big piece of white paper with a little target test design in the middle.
In single AF point, determine the area or location of the actual AF point. Then go to D9 AF. Do the test again but as the main AF point goes off the test image and onto the white BG, one of the AF points surrounding it should still hold it in focus until you move away from the area that point can hold.
In 3D mode. use the center AF point. As you pan away from the little test patter, the AF point will move and keep the test image in focus - it'll follow. English is my second language, but when I read the manual, that's what I gather. I don't know how or why the other bodies perform as people describe, but could it be that they aren't working properly as described?
That's actually part of the behaviour we've noticed.
In the wall target test, there is nothing for the primary AF point to lock onto after it's moved off the cross-hatch area, so it will stick. You can also de-fcous the lens and keep the primary AF point off of the cross hatch area and the outside sensors will pick it up. This seems to work as expected with Dynamic and is something I noticed about 6 weeks ago shooting gulls and testing. If they were against the sky (so the main AF point had nothing to lock on), then they would stick with the subject.
The thing is that in the older bodies (I use the D810 for an example, but this is true for all the the pre-D5/D500 bodies), Dynamic area would stay with a target even if what was under the primary AF point was AF friendly. The new system does not - and that's the issue. Like I say, if it's going to more or less act like single point, why not just shoot single point? What was the point of having it there in the first place?
Oh, and I know about 3D - and it works as it should on the D5/D500. I'm not really trying to figure out how to photograph anything, I'm trying to determine if everyone else is seeing similar behaviour in Dynamic to what I'm seeing to verify that something has changed.
Lance B wrote:
Actually, that is a good point. It does seem a large AF area for such a small subject and I wonder if this has some significance. Maybe try 72 or even 25 and see if it happens for this same subject size? Maybe the AF areas are more responsive the close the Single point and lose effectiveness the further you get away from the Single point?
I've tried it with all the Dynamic areas - same results at D153, D72, and D25.
I can confirm that same behavior on my D500 but I do not see it on my D750. Clearly there is an issue with Dynamic Area Mode on the D500. Have you tried contacting Nikon customer support to raise this issue?
Amit
Had a back and forth that went nowhere. They kept referring to D9 on the D500 - it only goes down to D25, so you can image how helpful they were. Oh, and when they couldn't answer the question, they suggested I try a different AF mode...Sigh
I think it's important to consider that the camera wants to use the selected primary AF point to focus in Dymanic AF. Those extra helpers are used briefly as the primary AF point tries to achieve focus. As soon as it can again, it does.
Steve Perry wrote:
I was going to ask the same as Snapsy above. Also, what was behind the primary AF point? In my tests, there needs to be an AF-friendly target behind the primary AF point or it will actually seem to stick with the main target. I'll post another example shortly.
Thanks!
Yes, there were branches behind the branch I was focused on for the camera to easily refocus on. I was also using 5fps and although ACR does not show a focus point, the shots were in focus. If I get a chance today. I'll go out in the yard with one of my dogs and see what happens. Even still, if this is "normal" behavior we are seeing, it is indeed very different than the 810 and any of the previous pro bodies (D3, D4) I have owned.
la puffin wrote:
I think it's important to consider that the camera wants to use the selected primary AF point to focus in Dymanic AF. Those extra helpers are used briefly as the primary AF point tries to achieve focus. As soon as it can again, it does.
Which is as it should be for Dynamic - the primary AF point always has priority when getting the initial lock, although the system will choose another area in the field if there's nothing for it to lock on.
What's different now is what happens when it loses the lock. In older bodies, it would hand off AF responsibilities to another AF point (and you can see this with the first post on this thread). Anyone using those bodies can shoot in Dynamic and when viewing the the AF point location could clearly see the camera switches AF points when it loses a lock. Additionally, the camera would hold that lock on the new point indefinitely, not release it after a short delay as we see now.
In the photos of my daughter running, keep in mind that the time from it having focus on her to focusing on the background was only 2 frames at 5FPS - so it's dropped the main target in less than 2/5ths of a second - not a lot of time for a photographer to react in a fast paced action scenario. Essentially, it's like shooting in single point mode.
Of course, it's possible that Nikon has changed the way this works and is only using the outer points to "enhance" tracking, but in doing so they taken away a valuable part of the system.
RKnecht wrote:
Yes, there were branches behind the branch I was focused on for the camera to easily refocus on. I was also using 5fps and although ACR does not show a focus point, the shots were in focus. If I get a chance today. I'll go out in the yard with one of my dogs and see what happens. Even still, if this is "normal" behavior we are seeing, it is indeed very different than the 810 and any of the previous pro bodies (D3, D4) I have owned.
If possible, try to have your dog run at the camera in d153, moving the AF point off of him as you do so - but make sure it's sitting over an AF friendly target. Set A3 to the default of "3" and in the middle of "erratic" and "stead" - basically the defaults. Note that center AF point may be a touch bigger than it appears in the viewfinder, so you'll need it well off the dog while keeping the pup well within the d153 area.
You might also try a left-to-right panning shot as well (if possible, not sure how cooperative your dog is!) I've done this with a runner, thinking that if the camera were moving as well as the subject it would work but in my tests, it drops the target.
Shooting Owls in the field camera will drop the target unless sticking up to the sky.
Using the D810 alongside D5/D500 its very noticable.
Crazy as this is going to seem, the Dynamic AF works better on the D810 than my D5/D500, thats the conclusion I came too some time ago and I tried to forget about it, until your post
Steve Perry wrote:
Anyone using those bodies can shoot in Dynamic and when viewing the the AF point location could clearly see the camera switches AF points when it loses a lock. Additionally, the camera would hold that lock on the new point indefinitely, not release it after a short delay as we see now.
Hi Steve, I'm just trying to make sense of all this, and I really appreciate all of the time you've put in helping others with your posts, videos and the delicious cookies you sent me for my birthday
Here's something to consider. Could this be a deliberate change of function that is directed at shooting sports? Remember, the D5 was released in time for the Olympics. Shooting sports, you want use the AF point you've chosen because most of the time, one is shooting wide open whether for separation from the background or because of shooting in low light conditions. With a razor thin DoF, accuracy with the AF point is crucial for the difference between a pretty sharp shot and a tack sharp shot. Dynamic AF is a helper to get a close relative until the photog can get the AF point back onto (a contrast) target, and as soon as that happens, the computer in the camera goes back to the primary point. This is what one would want - AF on the chosen point.
A couple of little scenarios:
1) D9 for basketball. Focus on a eye, player moves, AF point goes on a check and can't focus. Nearby point picks up contrast on the side of the nose, shot is made and within the range of DoF. Photog gets the primary AF point back on the eye and continues. Don't you want the AF back on target instead of staying with the helper? I want it to ficus where I'm trying to focus. Using D25, the larger area might catch his hand near his face as he's making a shot and DoF is too short and you get a sharp hand and slightly OOF face.
2) D25 for motosports. Shooting a black car with white number on the side. That's a nice big contrast point, but the photog loses the white number and the primary AF point moves off slightly and loses contrast to focus as all black is in the primary AF point. D9 is too small, but D25 can reach the side of a number and achieve AF. Being the crack shot that I am (ha ha ha), I move the primary AF point to the driver's head in the window. I don't want the camera still using the helper, I want it to go to the primary AF point. Doesn't this make sense? DoF is probably deep enough to cover it, but it's just an example.
Thinking further about the D5 as a sports centric camera, is the issue of lower DR at low ISOs. Shooting sports in daylight with a fast aperture, under exposing isn't really an issue, unless a clown like me has EC set down two stops. What's more likely is over exposing, and I can't remember where I read it, but the D5 has a good degree of headroom here (highlight recovery). Once the photog starts shooting in less light (and with a higher ISO) is where DR is more important and which is where the D5's DR is improved over other cameras.
I could be completely wrong, but I'm tossing out that there is another way at looking at the function of the camera. I'm not being a D5 apologist, I'm not going to dump it if they changed the AF behavior to be more like the D810. I also have too much spare time this morning if I'm thinking about these kind of things. I should go buy some shoes
la puffin wrote:
Hi Steve, I'm just trying to make sense of all this, and I really appreciate all of the time you've put in helping others with your posts, videos and the delicious cookies you sent me for my birthday
Here's something to consider. Could this be a deliberate change of function that is directed at shooting sports? Remember, the D5 was released in time for the Olympics. Shooting sports, you want use the AF point you've chosen because most of the time, one is shooting wide open whether for separation from the background or because of shooting in low light conditions. With a razor thin DoF, accuracy with the AF point is crucial for the difference between a pretty sharp shot and a tack sharp shot. Dynamic AF is a helper to get a close relative until the photog can get the AF point back onto (a contrast) target, and as soon as that happens, the computer in the camera goes back to the primary point. This is what one would want - AF on the chosen point.
A couple of little scenarios:
1) D9 for basketball. Focus on a eye, player moves, AF point goes on a check and can't focus. Nearby point picks up contrast on the side of the nose, shot is made and within the range of DoF. Photog gets the primary AF point back on the eye and continues. Don't you want the AF back on target instead of staying with the helper? I want it to ficus where I'm trying to focus. Using D25, the larger area might catch his hand near his face as he's making a shot and DoF is too short and you get a sharp hand and slightly OOF face.
2) D25 for motosports. Shooting a black car with white number on the side. That's a nice big contrast point, but the photog loses the white number and the primary AF point moves off slightly and loses contrast to focus as all black is in the primary AF point. D9 is too small, but D25 can reach the side of a number and achieve AF. Being the crack shot that I am (ha ha ha), I move the primary AF point to the driver's head in the window. I don't want the camera still using the helper, I want it to go to the primary AF point. Doesn't this make sense? DoF is probably deep enough to cover it, but it's just an example.
Thinking further about the D5 as a sports centric camera, is the issue of lower DR at low ISOs. Shooting sports in daylight with a fast aperture, under exposing isn't really an issue, unless a clown like me has EC set down two stops. What's more likely is over exposing, and I can't remember where I read it, but the D5 has a good degree of headroom here (highlight recovery). Once the photog starts shooting in less light (and with a higher ISO) is where DR is more important and which is where the D5's DR is improved over other cameras.
I could be completely wrong, but I'm tossing out that there is another way at looking at the function of the camera. I'm not being a D5 apologist, I'm not going to dump it if they changed the AF behavior to be more like the D810. I also have too much spare time this morning if I'm thinking about these kind of things. I should go buy some shoes ...Show more →
Per Steve's discovery, the dynamic AF on the D5/D500 will never select an AF point other than the initial AF point set by the user. In that respect it's no different than single-point AF-C. It's definitely broken.
I'm thinkin' it's true...ignorance IS bliss.
Using Group AF for Rodeo/BIF and it doesn't miss.
Just putting that out there for anyone ready to throw
their D500 (D5) under the bus. It just works as intended.
The way that Dynamic AF functions on the D5 (using AF C), seems to be that the lens focuses on the object that is immediately under the centre of the D-pattern; and the camera then holds focus on that item until that item passes beyond the outer edges of the focus-group "frame".
So, at D153 you have a big window frame while at D9 you are left with a very tight frame for retaining focus.
As soon as the focussed object leaves that "frame", AF will change its focus to pick the object which now lies under its centre-point.
For Steve's series of his daughter, it would probably have worked better at D153?
Shooting Owls in the field camera will drop the target unless sticking up to the sky.
Using the D810 alongside D5/D500 its very noticable.
Crazy as this is going to seem, the Dynamic AF works better on the D810 than my D5/D500, thats the conclusion I came too some time ago and I tried to forget about it, until your post
Yup, saw the same thing when I was playing with the D5/D500 in Dynamic mode in FL a few weeks back. The D810 tracks noticeably better in Dynamic (as does the D7200, D4, etc).
