Z8/9 Auto Focus (C) w/Sub Detection 🐦+👁️=🧐

[ajrmd]

Thinking we may have wandered off the ranch a bit. The “poser” was the subject detection box on the critter’s eye was closer to the camera in the reflection than the duck itself. The muscovy duck shown here is a mock up of what I encountered. This focus issue only seems to occur when the water is smooth with one bird in the frame. Steve’s suggestion to not use S/D and use a small focus area will solve this problem. BIF is a different topic…

Yes, we may have digressed a bit, though I'll post some images of the issues I've noticed with Osprey/Eagles diving in a bit. For now, these are the types of situations where I notice the subject detect will switch to the reflection. As I mentioned, I think that it is fooled by the brighter, nearer object. In my experience, either moving the reflection out of frame, quickly switching to another, smaller AF mode, or reacquiring AF (i.e. stop BBF and re-engage) seem to be the best strategies when faced with this.

You can only see EXIF info for this image if you are logged in.

 

[ajrmd]

That has been my experience with both Canon and Sony, but not Nikon. Usually with my R5 or a9ii the tracking stayed solid all of the way through the strike and depending on the amount of spray and/or how deeply the bird travelled under water the af system might lose af capture. What I experience with the Z8 and again it matters not whether I am in AA, WS, WL, Custom, etc. is that the AF tracks amazingly well through the dive sequence until the bird nears the water and then the af system wigs out through the sequence going oof before the strike usually ff’ing) and then it recovers several frames later after the bird emerges. It doesn’t appear to be grabbing on to splashes or particular specular highlights, rather it seems like the af system just loses it. Again, the bird hasn’t changed position drastically in the frame nor does there appear to be any apparent confounding factors. When I get home in a couple of days, I’ll wade through some sequences and see if I can post several to demonstrate my experiences.

I should note that my recent foray shooting spoonbills was spectacular and none of these issues manifested for the thousands of frames I took of them landing. The AF was spot on even against water for >95-98% of the shots. For some reason it’s different for Osprey and eagles traveling parallel to the plane of the sensor and striking. I’ve tried adjusting other af settings though they don’t appear to make a difference.

OK, my apologies for the divergence though I wanted to address the issue of loss of AF capture/integrity, or however one would like to describe it for diving birds who are moving roughly parallel (or tangential to the plane of the sensor). If there is enough interest, I would be happy move this to another thread. Since updating to FW 2.0, I've noticed that the Z8 AF has improved significantly. Specifically, it performs better for diving birds such as pelicans , terns, etc. who dive fairly vertically and splash into the water as opposed to eagles and osprey who move through a catch. For the former birds, the AF stays locked on through the dive and after the bird hits the water. In contrast for the later striking birds, I've noticed that AF tracking suddenly gets wonky as the bird nears the water; it loses AF capture for a few frames usually just before the strike and it continues to mis-focus for a half dozen to a dozen frames and then reacquires the AF sometime later in the sequence. This has been a frustrating issue as I have not experienced this with most of my Canon and Sony gear and it doesn't seem to be amenable to changing AF modes (AA, WL, WS, etc.) or AF settings such as AF priority, AF speed, tracking sensitivity, etc. - believe me when I say I've tried nearly every combination and as my example sequence shows, there aren't any apparent confounding factors such as another more contrasty object in the frame, splashes, etc. This phenomenon is repeatable and consistent, occurring about the same time in every strike.

Anyhow, I've been experimenting with stopping the continual AF by releasing the BB as the bird nears the water and while that can help as long as one has sufficient DOF, in real life at 20 FPS, it is more challenging that it seems. Nonetheless, I would be curious to hear how others are overcoming this apparent conundrum. Do you see this in the field and if so, how are you able to overcome this? Any thoughts would be appreciated.

As for the images, it is difficult to include huge sequences which contain 100-150+ images. These frames aren't cropped and are the actual size of the subject and AF point as displayed in Nikon NxStudio. For brevity, I started midway through the dive as the bird nears the water (labeled frame "one") and one can see the sequence proceeding in the filmstrip at the bottom. The preceding frames were in focus and tracking was working perfectly. The next frame, "two" the AF suddenly is "lost" and appears front focused as it does for the next 7-8 frames, missing the strike, though the AF recovers on image "nine" and sticks with the bird as it emerges and flies away for another 50-60 shots in the sequence.

You can only see EXIF info for this image if you are logged in.

You can only see EXIF info for this image if you are logged in.

You can only see EXIF info for this image if you are logged in.

You can only see EXIF info for this image if you are logged in.

You can only see EXIF info for this image if you are logged in.

 

[JoelKlein]

Which AF mode and size was it set?
What’s the reason it’s not catching the eye?

Sony, Canon are better at this specific challenge?

 

[ajrmd]

Which AF mode and size was it set?
What’s the reason it’s not catching the eye?

Sony, Canon are better at this specific challenge?

1) In this case it was AA, though that doesn't really matter as the same scenario occurs in all AF modes and settings.
2) Distance perhaps or alternatively the Z8 struggles at times with osprey and other birds (animals) in identifying the eye, instead latching on to a bright area, in this case the body.
3) Yes!

 

[JoelKlein]

1) In this case it was AA, though that doesn't really matter as the same scenario occurs in all AF modes and settings.
2) Distance perhaps or alternatively the Z8 struggles at times with osprey and other birds (animals) in identifying the eye, instead latching on to a bright area, in this case the body.
3) Yes!

Maybe I need a Sony for diving birds and Nikon for all other stuff

Mark Smith told me he is using a Z8-600pf combo for last couple months now. I will ask him how he gets around it

 

[BradleyD]

I don’t mind the subject detection area moving out of my focus area as I can use it as a guide to get back on focus. Maybe it will bother me more as my skills with the Z8 improve. The bigger issue for me is the system confusing water reflected subjects and bouncing around. Turning off subject detection works, sometimes recomposing works, but it has happened to me enough times that I logged into my Nikon account and reported it as an issue. Figure if Nikon gets enough complaints, the issue will make it to the top of the Product Manager’s backlog and can hopefully be addressed in one of their firmware updates. It’s an interesting challenge and Nikon’s commitment to continuous improvement with each release has been impressive.

 

[ajrmd]

Maybe I need a Sony for diving birds and Nikon for all other stuff

Mark Smith told me he is using a Z8-600pf combo for last couple months now. I will ask him how he gets around it

Diving birds which move vertically into the water such as pelicans are not an issue, rather the challenges are with birds which tend to swoop in parallel or tangential to the sensor, strike and continue such as osprey and/or eagles. As I mentioned, the phenomenon where the AF loses capture before the strike and regains it afterwards is repeatable and recurrent.

 

[abc123brian]

Maybe I need a Sony for diving birds and Nikon for all other stuff

Mark Smith told me he is using a Z8-600pf combo for last couple months now. I will ask him how he gets around it

I don't think that is needed at all. I am not a bird photographer by any means, only shoot them when I am not shooting other wildlife, but I am not seeing the issues being discussed here. Most of the shots that ajrmd posted show the camera focused on the water in front of the bird. I don't know what settings was used, he said auto area, which does allow the camera to focus on anything in the frame. I just looked back through a set of photos I had of a diving osprey that I turned last minute to say diving. I did not have a chance to change my settings but fired off some shots. Using AA like ajrmd did, I noticed the camera did jump to the nearest foreground when the bird was underwater with the exception of a wing. I can see why this would happen since the bird is no longer recognizable for subject detection and the camera defaulted to nearest object. But immediately upon emerging from the water, the camera jumped back to the Osprey. Also, if I had been intending to shoot the osprey, I would have adjusted the way I was shooting with a smaller box to prevent this. I believe wrong settings are the culprit here. Also, I do not believe that Sony or Canon is any different in this regard.

Keep in mind that Mark Smith specializes in this type of photography and has developed a skillset that allows him to capture these images with just about any camera. Remember when everyone was complaining about the original Z7 not being able to focus? Check out Mark's images that show perfectly sharp images of Osprey diving with it. My point is what is easy to him may be difficult for others. I did see some of his videos were sponsored by Sony so I hope he is still able to provide unbiased observations regarding other cameras.

This is the shot in my series where the camera lost focus. As you can see, auto area lost the subject and jumped to the nearest thing in the frame. I'm actually surprised it didn't jump to the gallinule

You can only see EXIF info for this image if you are logged in.

 

[ajrmd]

I don't think that is needed at all. I am not a bird photographer by any means, only shoot them when I am not shooting other wildlife, but I am not seeing the issues being discussed here. Most of the shots that ajrmd posted show the camera focused on the water in front of the bird. I don't know what settings was used, he said auto area, which does allow the camera to focus on anything in the frame. I just looked back through a set of photos I had of a diving osprey that I turned last minute to say diving. I did not have a chance to change my settings but fired off some shots. Using AA like ajrmd did, I noticed the camera did jump to the nearest foreground when the bird was underwater with the exception of a wing. I can see why this would happen since the bird is no longer recognizable for subject detection and the camera defaulted to nearest object. But immediately upon emerging from the water, the camera jumped back to the Osprey. Also, if I had been intending to shoot the osprey, I would have adjusted the way I was shooting with a smaller box to prevent this. I believe wrong settings are the culprit here. Also, I do not believe that Sony or Canon is any different in this regard.

Keep in mind that Mark Smith specializes in this type of photography and has developed a skillset that allows him to capture these images with just about any camera. Remember when everyone was complaining about the original Z7 not being able to focus? Check out Mark's images that show perfectly sharp images of Osprey diving with it. My point is what is easy to him may be difficult for others. I did see some of his videos were sponsored by Sony so I hope he is still able to provide unbiased observations regarding other cameras.

This is the shot in my series where the camera lost focus. As you can see, auto area lost the subject and jumped to the nearest thing in the frame. I'm actually surprised it didn't jump to the gallinule
View attachment 87654

With all due respect please read the original post and information provided. I am a WL photographer and have used the three major systems extensively for many years. This phenomenon does not occur routinely on Canon or Sony high end gear. As I indicated, the af loses captuRe as the bird is flying by the water whether one uses AA, WL, WS, etc. it is not an issue of the af point migrating to the water, splashes, or other subjects and it occurs always prior to the strike. It’s not attributable to difficulty panning, keeping the subject in the frame, distance to the subject, sudden illumination changes, etc. and again, this sample sequence would have been easily captured with an R3, R5, A1, A9(ii, iii). The loss of af does not occur with birds diving straight into the water, but those who strike. You posted one blurry image with a small subject at distance with many other confounding subjects. I am not surprised it was oof and your comments were not helpful.

 

[abc123brian]

With all due respect please read the original post and information provided. I am a WL photographer and have used the three major systems extensively for many years. This phenomenon does not occur routinely on Canon or Sony high end gear. As I indicated, the af loses captuRe as the bird is flying by the water whether one uses AA, WL, WS, etc. it is not an issue of the af point migrating to the water, splashes, or other subjects and it occurs always prior to the strike. It’s not attributable to difficulty panning, keeping the subject in the frame, distance to the subject, sudden illumination changes, etc. and again, this sample sequence would have been easily captured with an R3, R5, A1, A9(ii, iii). The loss of af does not occur with birds diving straight into the water, but those who strike. You posted one blurry image with a small subject at distance with many other confounding subjects. I am not surprised it was oof and your comments were not helpful.

I apologize my post offended you. The original post was regarding the camera focusing on the eye in the reflection in the water. I was responding to Joel's question which your 'helpful' answer is that Nikon is not capable of tracking a diving Osprey and you need a Canon or Sony, which seems to be the answer you're sticking to. I am sorry you found my response unhelpful, but hopefully a more open minded person could. My one blurry photo with a small subject shows that when the camera loses the subject, it focuses on the closest thing inside the focus box, in this instance, the camera was in AA.. If you read my comment, this photo is where the camera lost focus. If I am using a Wide Area mode (usually C1 with a smaller box, this doesn't happen. If you look at the series you posted you may notice your image is focused in the same place, not where the red box is placed in the image. I am not experiencing the same issue your are. Many others are not either so it is very possible it is something you have configured or doing differently.

There is a post from a user on DPReview who has a Z9 and A6700 who had posted many times that the A6700 was superior to the Z9. He recently performed a complete reset and reinstalled the firmware on the Z9 and reset from scratch. Here is a quote from his thread about it "I have NO idea what magic a complete reset accomplished, but do I ever feel blessed. This camera has literally been transformed into a tracking monster where before it was handicapped and my Sony A6700 ran rings around it." Maybe it is worth a try.

 

[JoelKlein]

I asked Mark, he confirmed every system needs intervention to stop the AF when it is going under.

I just looked through thousands of images me and my wife were bursting at 20fps on flying Ospreys. They really gave us a run for it flying near and far, until they splashed. We were both in Auto area. 98% is in focus. The other 2 percent is due to awkward positions of the bird.

Only when she tried using the 70-200+2xTC+DX, there are a couple out of focus before and after the splash.

BTW, NX Studio latest update delivered gorgeous post processing especially in clean noise reduction

Here are 2.
I’m going to open another thread in post processing

You can only see EXIF info for this image if you are logged in.

You can only see EXIF info for this image if you are logged in.

 

[ajrmd]

I asked Mark, he confirmed every system needs intervention to stop the AF when it is going under.

I just looked through thousands of images me and my wife were bursting at 20fps on flying Ospreys. They really gave us a run for it flying near and far, until they splashed. We were both in Auto area. 98% is in focus. The other 2 percent is due to awkward positions of the bird.

Only when she tried using the 70-200+2xTC+DX, there are a couple out of focus before and after the splash.

BTW, NX Studio latest update delivered gorgeous post processing especially in clean noise reduction

Here are 2.
I’m going to open another thread in post processing

Glad to hear you are obtaining good results and again, the issues I am encountering have nothing to do with splashes or other distractors such as water, reflection, etc. For some unknown reason, the Z8's AF tracking seems to break capture as the bird nears the strike, it continues oof for a few frames and then it re-acquires focus a few frames afterwards as illustrated in the example.

 

[JoelKlein]

Glad to hear you are obtaining good results and again, the issues I am encountering have nothing to do with splashes or other distractors such as water, reflection, etc. For some unknown reason, the Z8's AF tracking seems to break capture as the bird nears the strike, it continues oof for a few frames and then it re-acquires focus a few frames afterwards as illustrated in the example.

Maybe the Z8 does have an issue the Z9 doesn’t.

Those out of focus shots taken with a 70-200+2xTC+DX were shot with the Z8…

 

[JoelKlein]

Here is a GIF of 48 photos taken with the Z8 70-200+2xTC+DX

https://www.dropbox.com/scl/fi/u1ah564pxs3dx1y8n3xwm/Lightroom_6pQJT2QpX1.gif?rlkey=lxrvzy7c738940uo3x2zhl2zb&dl=1

 

[ajrmd]

Here is a GIF of 48 photos taken with the Z8 70-200+2xTC+DX

https://www.dropbox.com/scl/fi/u1ah564pxs3dx1y8n3xwm/Lightroom_6pQJT2QpX1.gif?rlkey=lxrvzy7c738940uo3x2zhl2zb&dl=1

I'm not all that surprised by the camera's behavior in that scenario (though it would be helpful to know the settings). The subject was small in the frame and was moving against a high contrast background. In those situations, and given the size of the subject, I would normally be in wide-small rather than AA. My experiences have been different as I am shooting against clean backgrounds with larger subjects. The camera has no problem following the bird down to the water and then just a couple of frames before the strike - again no high contrast distractions - the AF loses capture for a few frames and then regains composure a couple of frames afterwards.

 

[JoelKlein]

I'm not all that surprised by the camera's behavior in that scenario (though it would be helpful to know the settings). The subject was small in the frame and was moving against a high contrast background. In those situations, and given the size of the subject, I would normally be in wide-small rather than AA. My experiences have been different as I am shooting against clean backgrounds with larger subjects. The camera has no problem following the bird down to the water and then just a couple of frames before the strike - again no high contrast distractions - the AF loses capture for a few frames and then regains composure a couple of frames afterwards.

Setting was AA Bird AF.
Same all with Z8/9 600pf 800pf and no issues there. Although the bird never really submerged.

 

[EricBowles]

Seeing a number of posts here, I wonder how much of the issue is the camera seeking an alternate target because the subject is not recognized as a bird. Sometimes it's the angle of the bird, the head position, the coloring, or even the head being in shadow. If the subject is out of focus, sometimes the camera jumps to another out of focus target thinking that it can sharpen that image. The same is true for a subject partially in shadow - particularly when the head is in shadow. Comparisons across brands with well lit subjects against clean backgrounds is not a true comparison of the variables involved. Cameras look for contrasty targets and specific shapes, then start looking at head and eye. Head on views can be a challenge because the subject is hard to recognize.

 

[Steve]

Seeing a number of posts here, I wonder how much of the issue is the camera seeking an alternate target because the subject is not recognized as a bird. Sometimes it's the angle of the bird, the head position, the coloring, or even the head being in shadow. If the subject is out of focus, sometimes the camera jumps to another out of focus target thinking that it can sharpen that image. The same is true for a subject partially in shadow - particularly when the head is in shadow. Comparisons across brands with well lit subjects against clean backgrounds is not a true comparison of the variables involved. Cameras look for contrasty targets and specific shapes, then start looking at head and eye. Head on views can be a challenge because the subject is hard to recognize.

Couple that with the fact that many forget - once Subject Detection loses the subject, it reverts back to the AF area it was using before Subject Detection took over. So, if it was in Auto and loses the subject, you basically have Auto AF behavior. Same with 3D, Wide.

 

[Ken Miracle]

Depends - in theory it'll try to stick, and if the eye is distinctive enough it just might. However, it's using a different system based primary on color differences. I've found 3D without SD often sticks to the eye for a moment but when the subject moves, starts to wander.

Honestly, sometimes you just gotta use a single AF point and not mess around. It's nice when the tech helps, but it's still in its infancy (Subject Detection that is) and it's not perfect. I get a heck of a lot more keepers if I switch it off / use single point when it struggles.

I used single point for almost everything in my DSLR's and I have a lot of BIF on the wall taken using that. I did start using group on BIF when it came out in D4S and later bodies.

On Z9 I have used wide area custom 1x1 with a one button push to toggle subject detection off and on and this has worked well on birds 99% of the time when other AF Area modes were having a problem. In my birding world I have little time to fiddle it has to be fast or the bird is gone.

Closest thing to group I have found and it is my who knows what is going to pop up default af are mode most of the time is wide area custom 5x3. I have one button pushes set to instantly hand off to AF Area mode, 3D and wide area custom 1x1.

Now with my video record button set to toggle through af area modes quickly I can get to single point quickly and that still rocks.

 

[Ken Miracle]

Makes sense.
On BIF handholding, single point is impossible. Spoonbills, Ospreys, Eagles, etc.

I used single point for almost everything in my DSLR's and I have a lot of BIF on the wall and in use by non profits and on others walls taken using single point. I did start using group on BIF when it came out in D4S and later bodies. All those years of shooting shotguns probably helps.

On Z9 I have used wide area custom 1x1 with a one button push to toggle subject detection off and on and this has worked well on birds 99% of the time when other AF Area modes were having a problem. In my birding world I have little time to fiddle it has to be fast or the bird is gone.

Closest thing to group I have found is wide area custom 5x3.

Now with my video record button set to toggle through af area modes quickly I can get to single point quickly and that still rocks.

 

[Ken Miracle]

I’d put that down to Focus Tracking Lock On Delay. Hard to say how that was set in this case but if it’s set mid range or higher the AF system won’t immediately grab a new target at a different distance and will pretty much keep the lens focused at the same distance for a short time.

Yup ... why I was able to track BIF with DSLR's for years as they moved through trees, grasses etc. let alone diving Osprey's.

 

[Ken Miracle]

Like arbitrage said, getting lower is going to make a difference. Once the bird hits the water you’re going to have a lot of shiny, reflective water drops and splashing to potentially confuse the AF. A lot of it is also on the photographer to keep the bird in one place during the dive and not go completely past it when it hits the water making a sudden stop.

I have never gotten ahead of a bird in a dive and if you believe that I have a bridge in Brooklyn I can sell you at a great price Great point.

 

[Dinusaur]

Seeing a number of posts here, I wonder how much of the issue is the camera seeking an alternate target because the subject is not recognized as a bird. Sometimes it's the angle of the bird, the head position, the coloring, or even the head being in shadow. If the subject is out of focus, sometimes the camera jumps to another out of focus target thinking that it can sharpen that image. The same is true for a subject partially in shadow - particularly when the head is in shadow. Comparisons across brands with well lit subjects against clean backgrounds is not a true comparison of the variables involved. Cameras look for contrasty targets and specific shapes, then start looking at head and eye. Head on views can be a challenge because the subject is hard to recognize.

Yes that is most often the case for diving birds that are far away. For a close-up subject not sure why this happens sometimes. Here's an example of two consecutive full frame shots of a diving Osprey, the subject was close and nothing much changed in terms of light etc. yet, the first frame was in focus but not the next one - the focus shifted to the background for no apparent reason. I was using Z9 + 400tc, in Wide-L and Bird subject mode. Maybe it all comes down to shooting techniques, but when a large number of users is having similar problem, it generally indicates some sort of shortcomings in the camera itself.

You can only see EXIF info for this image if you are logged in.

You can only see EXIF info for this image if you are logged in.

 

[apakett]

Even when AF system evolve enough that they don't let go until the bird is underwater, a situation involving diving will still demand some skill on the part of the photographer.

Steve, can you explain the technique? Are you saying that between the time a bird hits the water, when the camera looses autofocus, and when the bird emerges with a fish, you change autofocus modes? I am lucky to keep the camera on the subject during that time.

 

[Judy Lapetino]

Hope you see this one Steve… I ran into an annoying issue with auto focus and subject detection. It finds the eye and nails the focus….in the reflection that is closer to the camera. This “yellowleg” is an example (mock-up) of what I occasionally encounter. The calmer the water the more frequently this occurs. The result is an out of focus bird if you shoot it anyway. I’ve been able to cope with this by re-framing the shot and that usually gets the eye detection back on the bird, not the closer reflection….
View attachment 87201

I had exactly the same issue with my Z8 when the reflection pond was still. I switched to spot focus.