Home · Register · Software · Software · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Nikon Forum | Join Upload & Sell

1       2      
3
       4              32       33       end
  

UPDATE - Nikon D5/D500 Dynamic AF Issue
  
 
Steve Perry
Offline
• • • • •
Upload & Sell: On
p.3 #1 · p.3 #1 · UPDATE - Nikon D5/D500 Dynamic AF Issue


snapsy wrote:
Thanks again. Question: For the case where you move the initial AF point (center) off the subject and onto something unfocusable (like the sky), does the camera properly use an adjacent AF point for focusing, as determined by the active AF point displayed in the image review of that photo?


That's one of the things I find concerning. It always seem to indicate the center AF point, but it still "locks on" even if the subject is not under the AF area (note it will only do this with a blank sky - and locking on to something else in the AF field is a normal behavior int his case). OTOH, the "older" bodies will show a different AF point.



Mar 01, 2017 at 06:34 PM
Steve Perry
Offline
• • • • •
Upload & Sell: On
p.3 #2 · p.3 #2 · UPDATE - Nikon D5/D500 Dynamic AF Issue


henry albert wrote:
And interestingly enough, in group mode it will not track.


That's odd. I've not had any trouble with Group and it always tracks fine. Accidentally in AF-S maybe? (I ask because I've done it myself)



Mar 01, 2017 at 06:36 PM
henry albert
Online
• • •
Upload & Sell: On
p.3 #3 · p.3 #3 · UPDATE - Nikon D5/D500 Dynamic AF Issue


Steve Perry wrote:
That's odd. I've not had any trouble with Group and it always tracks fine. Accidentally in AF-S maybe? (I ask because I've done it myself)


Yup. Suffered a brain fart.



Mar 01, 2017 at 06:41 PM
snapsy
Offline
• • • • •
Upload & Sell: On
p.3 #4 · p.3 #4 · UPDATE - Nikon D5/D500 Dynamic AF Issue


Steve Perry wrote:
That's one of the things I find concerning. It always seem to indicate the center AF point, but it still "locks on" even if the subject is not under the AF area (note it will only do this with a blank sky - and locking on to something else in the AF field is a normal behavior int his case). OTOH, the "older" bodies will show a different AF point.


Thanks. This indictes the problem is with the camera failing to implement the basic dynamic AF point handoff behavior vs it improperly acquiring a new tracked focus target.



Mar 01, 2017 at 06:44 PM
elkhornsun
Offline
• • • •
Upload & Sell: On
p.3 #5 · p.3 #5 · UPDATE - Nikon D5/D500 Dynamic AF Issue


With each new camera I have to test and see how the autofocus actually works as Steve has been doing. I was very disappointed with the D3 as it was much more inclined to backfocus than the D2x if there was a bright object in the background with a hard edge - the AF chose the easier subject on which to lock focus.

With the D750 that I bought because it had Group mode I found the autofocus to be no better than that of the D7200 with moving subjects or low contrast subjects (the system has to focus on an area of contrast and a "hard" edge).

With the D500 camera Nikon recommend 153 point mode for moving subjects. Evidently the dedicated autofocus processor can handle the increased amount of data and does better with more points to use for calculating distance and position of the subject.

With the D500 I find that Group, 3D, and manual focus work the most reliably. With a test setup like Steve's one is setting the camera up to fail with two white objects in the background and objects with hard edges for contrast detection. It is an interesting test but I do not think it is the best approach to evaluate the new autofocus system.

It would be interesting to test different modes of autofocus more empirically with a hummer feeding station.

I was frankly amazed at how well Group AF worked on the D500 when photographing very fast and erratically moving hummingbirds. it was dead on 100% of the time.




Mar 01, 2017 at 08:13 PM
Steve Perry
Offline
• • • • •
Upload & Sell: On
p.3 #6 · p.3 #6 · UPDATE - Nikon D5/D500 Dynamic AF Issue


elkhornsun wrote:
With each new camera I have to test and see how the autofocus actually works as Steve has been doing. I was very disappointed with the D3 as it was much more inclined to backfocus than the D2x if there was a bright object in the background with a hard edge - the AF chose the easier subject on which to lock focus.

With the D750 that I bought because it had Group mode I found the autofocus to be no better than that of the D7200 with moving subjects or low contrast subjects (the system has to focus on
...Show more

Keep in mind that the test I'm using isn't the only way I've tried this - I've seen the same thing happen in the field. Also, when I do the exact same test with the D810 and D7200, they pass with flying colors every time. I would think the system in the D5/D500 would be even better, not worse.



Mar 01, 2017 at 09:07 PM
louie champan
Offline
• • • • •
Upload & Sell: On
p.3 #7 · p.3 #7 · UPDATE - Nikon D5/D500 Dynamic AF Issue


I just got through doing the same test Steve with my D5 and D500 and they both worked almost identical to each other. When set to D72 or D153 they both lost focus before moving out of the AF sensor area, if the background had enough contrast in it. Did this test outside in the backyard with a 80-400mm lens set on 300mm and focusing on a subject 40-50 feet away. Set at D25 on the D5 hard to tell but I think it was acting pretty much the same way.


Mar 01, 2017 at 09:39 PM
kwilliam8
Offline
• • • •
Upload & Sell: On
p.3 #8 · p.3 #8 · UPDATE - Nikon D5/D500 Dynamic AF Issue


I need to make a correction to my earlier post on this thread. I did a large variety of tests in my backyard this afternoon with the D500 and the 300mm PF lens. In addition to the obvious and consistent problems with the larger D153 area, I now believe that the D72 and D25 areas are also messed up. However, sometimes they worked properly. For a particular situation (i.e., what was being framed), the results were consistent however. Generally, it seemed that the closer the background was, the more likely the focus shifted prematurely. The further away, the less likely. Contrast seemed to matter, but not always. And color differences might have come into play as well. In summary, unlike my previous post, I know believe that all the dynamic areas have problems (to one degree or another). As someone who uses these features regularly, I find this situation very disappointing! Thanks to Steve for looking into this more closely, and then posting this thread to get some additional data points.
Keith W.



Mar 01, 2017 at 11:49 PM
Steve Perry
Offline
• • • • •
Upload & Sell: On
p.3 #9 · p.3 #9 · UPDATE - Nikon D5/D500 Dynamic AF Issue


louie champan wrote:
I just got through doing the same test Steve with my D5 and D500 and they both worked almost identical to each other. When set to D72 or D153 they both lost focus before moving out of the AF sensor area, if the background had enough contrast in it. Did this test outside in the backyard with a 80-400mm lens set on 300mm and focusing on a subject 40-50 feet away. Set at D25 on the D5 hard to tell but I think it was acting pretty much the same way.


Thanks for taking the time to test, I appreciate it



Mar 02, 2017 at 12:33 AM
Steve Perry
Offline
• • • • •
Upload & Sell: On
p.3 #10 · p.3 #10 · UPDATE - Nikon D5/D500 Dynamic AF Issue


louie champan wrote:
I just got through doing the same test Steve with my D5 and D500 and they both worked almost identical to each other. When set to D72 or D153 they both lost focus before moving out of the AF sensor area, if the background had enough contrast in it. Did this test outside in the backyard with a 80-400mm lens set on 300mm and focusing on a subject 40-50 feet away. Set at D25 on the D5 hard to tell but I think it was acting pretty much the same way.


Thanks for the update



Mar 02, 2017 at 12:34 AM
 

Search in Used Dept. 



arbitrage
Online
• • • • • •
Upload & Sell: On
p.3 #11 · p.3 #11 · UPDATE - Nikon D5/D500 Dynamic AF Issue


I did a little more testing today and one thing I found that by turning up the setting for Blocked shot AF response to 5 makes the behaviour super obvious as it jumps to background things really fast when the AF point is barely off the intended subject. I had previously had it on 2 as per Steve's original D500 guides and turning it up to 5 just made the issue worse.

But in all my further testing today, one thing is for sure, the mode is next to useless and using 153 or 72 or even 25 seems really pointless based on the behaviour. I might as well just use centre point (or if I had a D5 I'd probably just use D9).

Anyways, for me Group AF is the really impressive AF mode but today I was shooting small Redpolls in trees and you just couldn't use Group AF as it would grab small twigs/branches around the bird (in the foreground). I used my D25 for this but Single would have been just as good. Still for most stuff and BIF, Group AF is the way to go so I'm not all that concerned with how these Dynamic modes are working for my use case. However, I do think Nikon should be looking into this especially if it used to work as described on older cameras like D810.



Mar 02, 2017 at 02:11 AM
Steve Perry
Offline
• • • • •
Upload & Sell: On
p.3 #12 · p.3 #12 · UPDATE - Nikon D5/D500 Dynamic AF Issue


arbitrage wrote:
I did a little more testing today and one thing I found that by turning up the setting for Blocked shot AF response to 5 makes the behaviour super obvious as it jumps to background things really fast when the AF point is barely off the intended subject. I had previously had it on 2 as per Steve's original D500 guides and turning it up to 5 just made the issue worse.

But in all my further testing today, one thing is for sure, the mode is next to useless and using 153 or 72 or even 25 seems really pointless
...Show more

Pretty much how I feel about it

The focus delay should not come into play at all unless the subject it totally out of the selected dynamic AF area, and yet it does. I too find I use Group or single point most of the time anyway, but there are times the larger coverage could come in handy.

Thanks for posting



Mar 02, 2017 at 02:18 AM
la puffin
Offline
• •
Upload & Sell: Off
p.3 #13 · p.3 #13 · UPDATE - Nikon D5/D500 Dynamic AF Issue


I tried it with my D5 and the camera performs as yours does, Steve.

However, (and I'm probably wrong) I read something that the Dynamic-area AF behaves differently. If the main focus point can't focus, then it uses information from the nearest 9, 25, 72 or 153 points to achieve AF.

For example, you're shooting basketball and the player has a big black beard. You focus on his eye, but he moves and you're now focused on that black beard and the main AF point can't focus on it. So, in D9, the camera uses info from the surrounding 9 points to achieve AF. A nearby point may catch the side of his noise or lips - something with suitable contrast, and achieve AF.

Isn't what you're talking about actually 3D focus mode? I don't have any other Nikon bodies besides a D5 and D500, so I can't see how the others perform.



Mar 02, 2017 at 02:39 AM
mp0363
Offline
• •
Upload & Sell: Off
p.3 #14 · p.3 #14 · UPDATE - Nikon D5/D500 Dynamic AF Issue


Steve Perry wrote:
Not really a question, just trying to verify a behavior discrepancy I'v noticed between the old and new AF systems in Dynamic mode. Note that every other AF function works as expected, it's only Dynamic mode that is showing the issue. Focus and recompose (single shot I assume) works as expected.


Thanks for clarifying that you are not pressing and releasing the focus to change your camera.

So (if I understand) in AFC you are pressing AFL and holding while moving the camera. Only 3D will hold that subject, otherwise the focus will move regardless of 153, 25, or 5. A focus that didn't change would be a broken camera.

Perhaps I still don't understand, but just now I played with my D500 and the only thing I noticed was a quicker response as I moved in from 153, limiting my focus points.



Mar 02, 2017 at 03:03 AM
mp0363
Offline
• •
Upload & Sell: Off
p.3 #15 · p.3 #15 · UPDATE - Nikon D5/D500 Dynamic AF Issue


arbitrage wrote:
...the mode is next to useless using 153 or 72 or even 25 seems really pointless based on the behaviour. I might as well just use centre point (or if I had a D5 I'd probably just use D9)...


I tried to get away with blind capturing on a roller coaster ride, so put it into d153 and held the camera backwards. It was a waste of time. Despite our face/body being proximal and large, the camera grabbed onto everything but.

I shoot a lot with d25 (back buttong AFC) since it's quick and easy for close shooting (20mm), but the winner for precision is still single point and 3D so I keep that on my front shutter when I really need the eyes and can afford the slight slow down in acquisition.

I understand that the birders like Grp and I may try that with soccer etc., when the time comes for it.



Mar 02, 2017 at 03:12 AM
arbitrage
Online
• • • • • •
Upload & Sell: On
p.3 #16 · p.3 #16 · UPDATE - Nikon D5/D500 Dynamic AF Issue


la puffin and mp0363 you both have brought up good points. Yes really only 3D mode is truly designed to remember what you first focused on and track it across all the points as you move or it moves.

I think you may be right that the Dynamic modes are using the points as assist points to focus if the main point doesn't have enough contrast to focus...but it doesn't work like 3D and remember what you were first focusing on but instead will focus on something else it finds more contrasty.

This is actually how Canon describes its AF 4 and 8 point expansion mode in that the surrounding points "assist" the centre point if it loses contrast but they aren't designed to remember what the original focus point was on.

If this is how Dynamic works and it doesn't remember the original subject then I really only see a use for the smallest of the Dynamic modes like D9 on the D5 (basically the same as Canon's modes) and maybe D25 is okay also. But I don't see much use of the D72 and D153 as how is the camera deciding what point to use if the centre point loses contrast? On the D500 it could be choosing a point off in the far corner of the frame in D153 mode. Seems like a strange mode to me....not sure how Nikon designed it??



Mar 02, 2017 at 03:25 AM
la puffin
Offline
• •
Upload & Sell: Off
p.3 #17 · p.3 #17 · UPDATE - Nikon D5/D500 Dynamic AF Issue


arbitrage wrote:
If this is how Dynamic works and it doesn't remember the original subject then I really only see a use for the smallest of the Dynamic modes like D9 on the D5 (basically the same as Canon's modes) and maybe D25 is okay also. But I don't see much use of the D72 and D153 as how is the camera deciding what point to use if the centre point loses contrast? On the D500 it could be choosing a point off in the far corner of the frame in D153 mode. Seems like a strange mode to me....not sure how
...Show more

D72 might be good for something larger or moving faster. Perhaps in a situation when you have to acquire a target quickly? I've never shot a bird, other than at other drivers, but I an erratic target might be useful with a larger area.

If Nikon stopped at D25, then some people would complain it's not big enough of an area. You know how that goes



Mar 02, 2017 at 03:43 AM
Steve Perry
Offline
• • • • •
Upload & Sell: On
p.3 #18 · p.3 #18 · UPDATE - Nikon D5/D500 Dynamic AF Issue


la puffin wrote:
I tried it with my D5 and the camera performs as yours does, Steve.

However, (and I'm probably wrong) I read something that the Dynamic-area AF behaves differently. If the main focus point can't focus, then it uses information from the nearest 9, 25, 72 or 153 points to achieve AF.

For example, you're shooting basketball and the player has a big black beard. You focus on his eye, but he moves and you're now focused on that black beard and the main AF point can't focus on it. So, in D9, the camera uses info from the surrounding 9 points
...Show more

3D and Dynamic are siblings, but not twins They perform similar functions, but in a different way. Your description of dynamic is pretty accurate - it starts with the main AF point (although it will pick another if it can't get an initial lock under the primary), and will hold on with that point as you track the subject. If the main AF point can no longer get a lock, it hands off AF responsibility to a nearby point (this is the part that's not happening anymore). As long as the subject stays in the field, it will continue to focus and track. The nice thing with Dynamic is that you can isolate the area to a much smaller space than 3D, thereby limiting the the choices the system has and isolating your focus.

3D also starts with a single point and will attempt to follow the subject around the entire AF area BUT is more more apt to jump points than Dynamic (this is the biggest difference IMO) - it will often jump even if it has a good lock to something that looks similar but is a touch more contrasty. Dynamic tends to be more "sticky" with the main AF point and subsequent AF points. 3D is also partially based on color, and tends to wander off target much more easily if the target is similar to the background, where Dynamic seems better in this regard (although Group is my preferred AF choice for tracking subjects on tricky backgrounds ). However, if you get a good subject that's a different color from the background, 3D can be very effective.

All that said..

Keep in mind that the behaviour has changed from the D810 and prior cameras. I can do these same tests with the equivalent settings in a D810 and D7200 and it works as expected. The D5/D500 do not. It's possible that Nikon decided to change them for whatever reason, but I don't see the point. They way they work now, they are little better than single point AF.

Anyhow, explanations aside, thanks for checking



Mar 02, 2017 at 04:17 AM
la puffin
Offline
• •
Upload & Sell: Off
p.3 #19 · p.3 #19 · UPDATE - Nikon D5/D500 Dynamic AF Issue


Steve Perry wrote:
Dynamic tends to be more "sticky" with the main AF point and subsequent AF points. 3D is also partially based on color, and tends to wander off target much more easily if the target is similar to the background, where Dynamic seems better in this regard

<snip>

Keep in mind that the behaviour has changed from the D810 and prior cameras. I can do these same tests with the equivalent settings in a D810 and D7200 and it works as expected. The D5/D500 do not. It's possible that Nikon decided to change them for whatever reason, but I don't see the
...Show more

Perhaps Dynamic works better with some types of subject matter. I mainly shoot sports and get a higher keeper rate using D9 than single point. 3D doesn't work so well for me in a number of situations, because as you described, the point can jump wildly if something of greater contrast enters the frame. As an example, that can sometimes easily happen in a gymnasium.

3D does work well in other types of shooting for me, but I have a hard time letting the camera decide what to focus on. I feel the same way about Auto AF. Auto ISO - that's different and I use it and shoot in Manual 99% of time. I'll let the camera ride that for me.



Mar 02, 2017 at 07:11 AM
Steve Perry
Offline
• • • • •
Upload & Sell: On
p.3 #20 · p.3 #20 · UPDATE - Nikon D5/D500 Dynamic AF Issue


la puffin wrote:
Perhaps Dynamic works better with some types of subject matter. I mainly shoot sports and get a higher keeper rate using D9 than single point. 3D doesn't work so well for me in a number of situations, because as you described, the point can jump wildly if something of greater contrast enters the frame. As an example, that can sometimes easily happen in a gymnasium.

3D does work well in other types of shooting for me, but I have a hard time letting the camera decide what to focus on. I feel the same way about Auto AF. Auto ISO
...Show more

That's the trick with Nikon's AF modes 3D, Group, and Dynamic have a lot of overlap and in many cases - well over half - you will get equally good results with any of them. However, under the right conditions, there are certainly advantages to one mode over the other.



Mar 02, 2017 at 01:19 PM
1       2      
3
       4              32       33       end






FM Forums | Nikon Forum | Join Upload & Sell

1       2      
3
       4              32       33       end
    
 

You are not logged in. Login or Register

Username     Reset password