Z9 Tracking Question

If you would like to post, you'll need to register. Note that if you have a BCG store account, you'll need a new, separate account here (we keep the two sites separate for security purposes).

Maljo

Well-known member
Subject Detection:
If 3D tracking puts its focus box on the eye of a moving animal is the subject tracking any different than if Auto Area AF finds the subject's eye and puts the focus box on the animal's eye? Once the eye is identified is the tracking any different from that point forward? In other words why switch from AutoArea AF to 3D tracking if AutoArea AF has found the eye?
 
Jan Wegener discussed that in the video linked here.


 
there have been other follow ups discussing using W-Area etc with 3D, and Lock-On setting(s) etc:


 
Not really any different once they are locked on.

Once subject detection takes over in either mode, it seems to work exactly the same. The big difference is that with 3D YOU pick the target. The problem in auto is that sometimes there's a subject that the camera can detect, but that subject isn't in focus. Unless the camera can "see" the subject, subject detection isn't going to kick in and the system will work normally until it does. This means that it's completely possible in Auto AF for you to have a subject that the camera would recognize in the frame but have the camera focus on a nearby tree or post because that subject was indistinguishable to the camera. Of course, the disadvantage with 3D is that if the subject is moving fast, it's sometime hard to pick it up.

OTOH, Auto is good when you need to pick up a subject quickly, or when the focus distance is about the same as the subject. There is a bit over overlap between them.

For 3D and action, I tend to use Wide (L) and when I see subject detection has identified the subject, I press Fn1 which is seton 3D on my camera and 3D takes over. This fixes both issues. It gives me a larger area for getting on the target than 3D, but I can also pick my target (unlike Auto).
 
The one thing I'd add is that when 3D does not have a recognized subject, it uses color as one of the main ways to identify an object. An example I saw recently was focusing with 3D tracking on a lime green tennis ball rather than the dog's eye because the color of the ball was easier to track.

One approach I've seen is to start with 3D but have a function button for Wide Large or Single. Normally with a clean background 3D will pick the right subject, but if it doesn't you can quickly refine and find a target, then let the camera take over. The assumption is that 3D is so good for your subject it will pick it up most of the time.

This is similar to what Steve described - but the opposite way of getting to the same point. Both approaches work depending on how easy focus is to acquire.
 
Further to earlier discussions, this thread particularly, i just tested yet again the impact of Blocked Shot Response in AF Lock On (a3) with W-Area on a terrestrial birds. With subject-detection enabled, if the focus box is moved off the subject or the subject moves, the AF cursor will stick on the subject for approx 1 sec on max delay '5' but it drops the subject almost instantly if set to '1'. I'm keeping it set on '5' as almost always focusing on a single subject (ie not moving among clustering subjects).

Nikon's official AF Tech Guide for sports does not give any settings for a3 with the 3D nor AutoArea modes [see Table pg 17]
 
Further to earlier discussions, this thread particularly, i just tested yet again the impact of Blocked Shot Response in AF Lock On (a3) with W-Area on a terrestrial birds. With subject-detection enabled, if the focus box is moved off the subject or the subject moves, the AF cursor will stick on the subject for approx 1 sec on max delay '5' but it drops the subject almost instantly if set to '1'. I'm keeping it set on '5' as almost always focusing on a single subject (ie not moving among clustering subjects).

Nikon's official AF Tech Guide for sports does not give any settings for a3 with the 3D nor AutoArea modes [see Table pg 17]
I find it very disappointing - and sad - that the Nikon Sports guide only addresses individual sports ((n)) other than soccer. Doesn't sound like they have much confidence in the Z9's ability when shooting basketball, football, flocks of birds, etc, etc!
 
I must be cynical as I learnt years back that Nikon's manuals have a weak relationship with reality ;) :) especially the deeper technical details and diverse challenges. However some of those sports throw up big challenges for any AF system: including gymnastics, BMX bikes, table tennis, and there's plenty evidence the Z9 delivers well on Rugby

I find it very disappointing - and sad - that the Nikon Sports guide only addresses individual sports ((n)) other than soccer. Doesn't sound like they have much confidence in the Z9's ability when shooting basketball, football, flocks of birds, etc, etc!
 
The Nikon Z9 AF road show just starting to be delivered through dealers does specifically address a large number of scenarios including sports like football. They share specific settings for AF Area, Blocked Shot Response, etc. along with the logic of why they chose those settings. I'll be hosting it through Nikonians in another 4-6 weeks. Nikon does not allow this program to be recorded so you will need to see it live.
 
My question is related to this but arises from my coming from a D850. 3D on the D850 was good for when the subject was moving back and forth from the camera (hence the "three dimensional" name!). A lot of my pictures arise where I'm panning with a subject that is moving parallel to the camera and I found other modes better. For the Z9 it seems that distinction now matters less -- am I getting that correct? In other words, even for subjects moving parallel to the camera that I am following with a pan, 3D seems to be the way to go? And, fwiw, my Z9 (ordered from B&H at 10:30 on 10/28) just arrived today! But even with it in my hands -- and not set up to my specifications yet -- this question about using 3D is at the front of my mind. Thanks for any thoughts!
 
During my couple weeks of Z9 ownership/testing I went back and forth trying to see what differences there was in using Auto vs 3D for my secondary AF mode to take over from Wide Area L (or S) (Steve's method above). I also went back and forth trying to test Auto vs 3D as the starting mode like what Eric mentioned above.

As a secondary mode I didn't find much between them as I only switched over if the subject had identified and then it seemed to do the same thing. As a starting mode I found strengths and weaknesses to both. As discussed above, 3D allows you control of where to start, Auto not at all. But I found Auto would often get distracted or not start where I wanted. But then I found 3D would sometimes just give up and jump to other things (often quite distant OOF backgrounds) if the subject was smaller in the frame.

If I went back again to test the Z9 I would certainly want to do more testing of Wide Area and Auto with Subject Detect Off. I don't think I tested that much (if at all) but I read now that people are finding that useful for certain situations and maybe that would have had them perform more like my favorite DSLR modes Zone and Auto??
 
My question is related to this but arises from my coming from a D850. 3D on the D850 was good for when the subject was moving back and forth from the camera (hence the "three dimensional" name!). A lot of my pictures arise where I'm panning with a subject that is moving parallel to the camera and I found other modes better. For the Z9 it seems that distinction now matters less -- am I getting that correct? In other words, even for subjects moving parallel to the camera that I am following with a pan, 3D seems to be the way to go? And, fwiw, my Z9 (ordered from B&H at 10:30 on 10/28) just arrived today! But even with it in my hands -- and not set up to my specifications yet -- this question about using 3D is at the front of my mind. Thanks for any thoughts!
I'd be thinking about Wide with Subject Detect activated for the scenario you are describing - at least to the point of acquiring the subject. 3D should work, but the key is you need to acquire focus first. If the subject is easy to pick up, that might work well with 3D, but you can also use other modes that limit the area such as Wide Large or Wide Small. In some cases Wide Large will be more reliable because you are constraining the area where the subject could be located.
 
Once subject detection takes over in either mode, it seems to work exactly the same. The big difference is that with 3D YOU pick the target. The problem in auto is that sometimes there's a subject that the camera can detect, but that subject isn't in focus.
My VERY PRELIMINARY tests tends to indicate, that Auto Area is more aggressive in searching for a recognizable subject than any other mode. It continues to search for the eye even when other modes default to head or torso. It seems to start searching at the center of the viewfinder.

These are my observations:
  • Recognition of the subject tries to detect eye -> head -> torso
    • Eye detection for a long neck animals (e.g. herons) tends to skip from eye/head to torso when the subject changes position and is reluctant to return back to the eye again
    • For small subjects ear or any other contrast object can be mistaken for an eye
  • Wide Area modes are more prone to sticking to anything evaluated as a subject (eye/head/torso) than Auto Area mode
  • Auto Area is less fixed on acquired subject, but tends to keep it even in a group of similar subjects if the changes of perspective are gradual
  • 3D tracking is better on staying fixed at particular subject feature (eye/head/torso), but it is more reluctant to acquire one
  • Main problem observed so far is the inability of the operator to force the camera to be more focused on the eye detection
My tests so far have been using birds as subjects. Single birds, groups of gulls and nesting herons. AF for other subjects may behave differently. I do not have enough evidence to be sure about the above points.
 
My VERY PRELIMINARY tests tends to indicate, that Auto Area is more aggressive in searching for a recognizable subject than any other mode. It continues to search for the eye even when other modes default to head or torso. It seems to start searching at the center of the viewfinder.

These are my observations:
  • Recognition of the subject tries to detect eye -> head -> torso
    • Eye detection for a long neck animals (e.g. herons) tends to skip from eye/head to torso when the subject changes position and is reluctant to return back to the eye again
    • For small subjects ear or any other contrast object can be mistaken for an eye
  • Wide Area modes are more prone to sticking to anything evaluated as a subject (eye/head/torso) than Auto Area mode
  • Auto Area is less fixed on acquired subject, but tends to keep it even in a group of similar subjects if the changes of perspective are gradual
  • 3D tracking is better on staying fixed at particular subject feature (eye/head/torso), but it is more reluctant to acquire one
  • Main problem observed so far is the inability of the operator to force the camera to be more focused on the eye detection
My tests so far have been using birds as subjects. Single birds, groups of gulls and nesting herons. AF for other subjects may behave differently. I do not have enough evidence to be sure about the above points.
All that sounds accurate with one exception. With Face/Eye AF the camera starts large and then narrows the search within that area. So it starts by looking for a person and the torso, then refines to the face, and finally the eye. If it loses the eye, it tries to stay on the face, and if that fails falls back to the body. That's why you still get sharp images even if the Eye AF box does not light up.

As you describe, the Wide modes are similar in that it looks for a torso or face in the Wide box, and refines to find the face then the eye. Since it does recognize the subject - person, animal, etc. - it also goes outside the borders of the box if necessary to find the face or eye of the subject.

This was the description by Mark Cruz of Nikon.

Alex Rae of Nikon indicated he uses all of his Fn buttons for specific AF modes. He starts with 3D and then can quickly refine if needed. As you point out, once AF is acquired, it tends to hold pretty well, so after acquisition he releases the Fn button and returns to 3D.

There is a lot of flexibility depending on what and how you are using the camera.
 
The Nikon Z9 AF road show just starting to be delivered through dealers does specifically address a large number of scenarios including sports like football. They share specific settings for AF Area, Blocked Shot Response, etc. along with the logic of why they chose those settings. I'll be hosting it through Nikonians in another 4-6 weeks. Nikon does not allow this program to be recorded so you will need to see it live.

Eric - where would one find this on Nikonians? I took a quick look but didn't see it. I'd be interested in watching.

Fred
 
Eric - where would one find this on Nikonians? I took a quick look but didn't see it. I'd be interested in watching.

Fred
We were not allowed to record the Button Tour program. I started the recording and was asked to stop. It is being delivered in a number of dealers.

Nikon NPS previewed the Z9 AF Program for NPS members only. It's being provided this week through Paul's Photo and will be available through a number of dealers. There are three programs as part of an ongoing series. The programs so far are Z9 Button Tour, Z9 Menus, and Z9 Auto Focus - released in that order. When we had the Button Tour program on March 8, the AF program had not been completed. We're planning to host all three programs on Nikonians with the first already completed.
 
I have watched all three NPS programs: Button tour, Menus and AF. I didn't find them any more helpful than reading the manual....... YMMV. The AF session was just too general and the speakers just showed some of their images without discussing AF modes used and why. The discussion never got close to how to set up the camera to switch modes quickly, etc.

I DO find discussions like this to be much more beneficial.
 
yah, i was wondering how valuable they were. certainly Eric's note about "along with the logic of why they chose those settings" got my interest.
They give a general description of each focus mode - but it is not much more than the manual descriptions. And while they may mention in passing what mode they used, that is about all that is said. I gave them feedback about the lack of details, etc. so maybe the next AF session they conduct will be better.

They also mentioned that they are working on ANOTHER video session that will be much more specific about focus modes, usage, and customizing buttons, etc. So I am cautiously hopeful about the FOURTH NPS session.....
 
I have watched all three NPS programs: Button tour, Menus and AF. I didn't find them any more helpful than reading the manual....... YMMV. The AF session was just too general and the speakers just showed some of their images without discussing AF modes used and why. The discussion never got close to how to set up the camera to switch modes quickly, etc.

I DO find discussions like this to be much more beneficial.

It was toward the end as Alex Rae was going through a series of images - each with the settings he used. On the Nikonians program we got into a bit more discussion about that issue over the last 20 minutes. We had a similar discussion with Jason Odell of the Image Doctors and in the roundtable programs. It's still a work in progress for Alex. I'd say using all the Fn buttons for AF settings is not really optimal - probably a waste of real estate. But it does provide a potential solution.

The images they showed at the end involved discussion for about a minute or so each. Each image had the AF settings they suggested along with commentary from the presenter about why those settings were recommended.

These programs are a bit like a firehose with the amount of information covered. 80% of it would be information you already know through discussions here. But there are also some nuggets, and some places where the speakers have different perspectives. For example, some speakers have advocated using the Shutter Release for focus, but the practical side for most pros is that there is so much muscle memory with AF-On that it is extremely hard to switch back to the shutter release. It is even harder if you are using another camera side by side.

All this is why the learning curve has been tough for the Z9. In many cases the answer is based on what you photograph.
 
Back
Top