RoamingScott wrote:
I tried it with a regular single point as well and got the same behavior, minus the flickering AF box. I was AFing the whole time in both clips. If Nikon understood that a green focus acquisition box in AF-C is important, you'd be able to tell
Thanks, that is interesting. Especially because some people, I think Thom Hogan said this aswell, claim that the Nikon Z AF is actually pretty good in practice.
That now clashes a fair bit with your very straightforward test that makes the Z AF look like a bit of a trainwreck, to be honest.
Not sure what to think.
Theodoric wrote: RoamingScott wroteI tried it with a regular single point as well and got the same behavior, minus the flickering AF box. I was AFing the whole time in both clips. If Nikon understood that a green focus acquisition box in AF-C is important, you'd be able to tell Thanks, that is interesting. Especially because some people, I think Thom Hogan said this aswell, claim that the Nikon Z AF is actually pretty good in practice.
That now clashes a fair bit with your very straightforward test that makes the Z AF look like a bit of a trainwreck, to be honest.
Not sure what to think.
In my testing so far, I'd say that AF-S is quick, decisive, and can be easily trusted...where AF-C is, as you say, an utter trainwreck that makes one shake their head in disbelief should you accidentally stop tracking the subject with pinpoint precision.
Now...huge grain of salt. I have NO IDEA what the consensus is on this 85/1.8 that I'm testing with. I know some lenses have quirks, and depending on their focusing motors, might struggle in certain circumstances. Until I have more lenses, and higher quality lenses to test with, I'm reserving my opinion to the experience of my first hand testing.
Edit: repeated this test with the 50/1.8 and got the same exact result.
lukemeup wrote:
And a 2nd test with A3 set to 'quick', while holding the focus button. Trainwreck?
Your background has a lot more light and contrast. The differences are obvious. Either way, I provided a control with my Sony, and it's obvious the Nikon AF engine isn't as robust.
Note that I won't be spending my day on my porch providing the forum with example videos. I can conclude, thus far in my testing, that the Nikon is more than a little behind the competition in AF-C for real world scenarios (not just test conditions with wildly varying contrasts).
I need the 100-400 in hand to really do some testing.
lukemeup wrote:
I'm not going to spend time shooting videos of a dark hedge either. Just wanted to show that Nikon's AF is not as bad as some 'tests' show it to be.
I can go outside and test it against a well lit foreground and background too. That's ideal conditions, and I'm sure it would perform fine. We don't often shoot in ideal conditions, and it's interesting to see when and how the AF system breaks down.
This might also be one of those moments where I gave you the tip of "use some manual focus" to nudge the camera along. There might be enough actually in the background for the camera to pick up on after that, it's kind of just lost in bokeh. I've had the camera do something similar to that going from close to infinity and vice versa a bit. It's all kind of just learning curve stuff since there really are just nuances to everything. RoamingScott wrote: lukemeup wroteI'm not going to spend time shooting videos of a dark hedge either. Just wanted to show that Nikon's AF is not as bad as some 'tests' show it to be.I can go outside and test it against a well lit foreground and background too. That's ideal conditions, and I'm sure it would perform fine. We don't often shoot in ideal conditions, and it's interesting to see when and how the AF system breaks down.
JadedWriter wrote:
This might also be one of those moments where I gave you the tip of "use some manual focus" to nudge the camera along. There might be enough actually in the background for the camera to pick up on after that, it's kind of just lost in bokeh.
I don't think being "lost in bokeh" should be a thing for a camera of this price point. If there is no contrast under the focus point, why does it freeze there? It should try to find focus instead, that's kinda the essence of AF, in my opinion.
Another test changing A3 to quick. Areas of bright/high contrast react faster, the can reacts faster, areas of shadow still refuse to lock on. I find that kind of interesting, given that one of Nikon's claims to fame is how well AF works in dark environments. You'd think that shadowy areas would sort of be the same thing.
It's not about the price bracket it's just the technology that's being used, since I've seen it happen in a few of my lenses on my Z6II. I basically get passed that by manual focusing it a little bit to get it to see something, but this is mostly a thing that I run into photographing people randomly when it's going from close to infinity and vice versa. Even if you watch his video you see it doing a better job focusing on the brighter parts of the background, not registering the dark parts and still just being confused about what to focus on when the background is a blur. Theodoric wrote: JadedWriter wroteThis might also be one of those moments where I gave you the tip of "use some manual focus" to nudge the camera along. There might be enough actually in the background for the camera to pick up on after that, it's kind of just lost in bokeh.I don't think being "lost in bokeh" should be a thing for a camera of this price point. If there is no contrast under the focus point, why does it freeze there? It should try to find focus instead, that's kinda the essence of AF, in my opinion.
Theodoric wrote: JadedWriter wroteThis might also be one of those moments where I gave you the tip of "use some manual focus" to nudge the camera along. There might be enough actually in the background for the camera to pick up on after that, it's kind of just lost in bokeh.I don't think being "lost in bokeh" should be a thing for a camera of this price point. If there is no contrast under the focus point, why does it freeze there? It should try to find focus instead, that's kinda the essence of AF, in my opinion.
I redid this test using single point but with AF-S, where I'd re-initiate focus in those same spots manually.
Guess what?
It worked 100% of the time, quickly, and without hunting/breathing.
The AF is capable...I just can't explain the discrepancy between AF-C and AF-S behavior.
Yeah I wasn't exactly joking when I told you, "Don't let the camera do too much thinking for itself." The camera works, it works fine but the system basically requires the person behind the camera to be active and basically pay attention to what the camera is doing and to actually be able to just know when to take over for it. Like I'm not typing hyperbole, it's the opinion of a person that actively shoots the system and has been doing so for almost a year. I'm crazy, but I'm not dumb. I'm hoping this stuff gets ironed out in some sort of FW update. The meat is there, but they need a better skeletal structure to house and support said meat. RoamingScott wrote: Theodoric wroteJadedWriter wroteThis might also be one of those moments where I gave you the tip of "use some manual focus" to nudge the camera along. There might be enough actually in the background for the camera to pick up on after that, it's kind of just lost in bokeh.I don't think being "lost in bokeh" should be a thing for a camera of this price point. If there is no contrast under the focus point, why does it freeze there? It should try to find focus instead, that's kinda the essence of AF, in my opinion.I redid this test using single point but with AF-S, where I'd re-initiate focus in those same spots manually.
Guess what?
It worked 100% of the time, quickly, and without hunting/breathing.
The AF is capable...I just can't explain the discrepancy between AF-C and AF-S behavior.
I appreciate your input everyone, many thanks for the video tests.
Thing is, I'm really eying the MC 105mm but I'm afraid that the Z AF will be disappointing. This thread didn't really help with that so far, I'm afraid
Using the a7R4 with the Sigma 105mm DG DN right now, which isn't even a great AF experience to start with, to be honest.
I really don't want to lug around the Z9 just to get acceptable AF...
What exactly would you be photographing that you think the MC105 wouldn't be able to handle it? Theodoric wrote:
I appreciate your input everyone, many thanks for the video tests.
Thing is, I'm really eying the MC 105mm but I'm afraid that the Z AF will be disappointing. This thread didn't really help with that so far, I'm afraid
Using the a7R4 with the Sigma 105mm DG DN right now, which isn't even a great AF experience to start with, to be honest.
I really don't want to lug around the Z9 just to get acceptable AF...
RoamingScott wrote: Theodoric wroteCan you compare AF performance in AF-C and single point?Caveat: the only similar lenses I have right now are the 85/1.8 on Z and 90/2.8 macro on FE.
I did this AF-C test at 2.8 on both, with both set to highlight meter.
Nikon failed to acquire the background in both dynamic and single point, and had this odd glitchy behavior in dynamic:
Sony paused a bit after losing the initial focus point but quickly found the background every time.
In a broader sense, I use wide mode (area equivalent on Nikon) all the time, with great results on Sony. The camera is sticky and can easily pick out obvious subjects. Nikon's area mode is one step above non-functional compared to Sony. I can see no use case so far where I'd pick area vs a single point.
If you can manage to keep the very small single point box, or slightly bigger dynamic box on the subject on Nikon, it seems to hold its own. I'd like to see more options on Nikon, like Sony's "center" or "zone", or at least be able to change the size of the single point box, like Sony's 3 flexible spot size options.
Single point seems roughly about the same (the lens will play a large role here) on these 2 lenses, with less focus breathing/settling on Nikon. The Nikon seems more sure of itself in AF-S than Sony, which is no surprise...Sony's AF-S mode is a no-go zone for many shooters because of its unreliability.
It really does feel like you're trading one mode for the other between these two bodies. ...Show more →
I believe the dynamic focus area extends beyond where the dots end so you likely still had the focus point on the can? That was weird.
Edit: I see same issue with single point. Probably lack of contrast in dark areas. I have a Z6ii and haven’t yet experienced that issue however the Z6 always struggled obtaining focus on objects out of focus, maybe the Z6ii’s slightly better low light capabilities help.
Buckeye2604 wrote: RoamingScott wroteTheodoric wroteCan you compare AF performance in AF-C and single point?Caveat: the only similar lenses I have right now are the 85/1.8 on Z and 90/2.8 macro on FE.
I did this AF-C test at 2.8 on both, with both set to highlight meter.
Nikon failed to acquire the background in both dynamic and single point, and had this odd glitchy behavior in dynamic:
Sony paused a bit after losing the initial focus point but quickly found the background every time.
In a broader sense, I use wide mode (area equivalent on Nikon) all the time, with great results on Sony. The camera is sticky and can easily pick out obvious subjects. Nikon's area mode is one step above non-functional compared to Sony. I can see no use case so far where I'd pick area vs a single point.
If you can manage to keep the very small single point box, or slightly bigger dynamic box on the subject on Nikon, it seems to hold its own. I'd like to see more options on Nikon, like Sony's "center" or "zone", or at least be able to change the size of the single point box, like Sony's 3 flexible spot size options.
Single point seems roughly about the same (the lens will play a large role here) on these 2 lenses, with less focus breathing/settling on Nikon. The Nikon seems more sure of itself in AF-S than Sony, which is no surprise...Sony's AF-S mode is a no-go zone for many shooters because of its unreliability.
It really does feel like you're trading one mode for the other between these two bodies.I believe the dynamic focus area extends beyond where the dots end so you likely still had the focus point on the can? That was weird.
Check my other vids, does the same thing with single point.
RoamingScott wrote: Buckeye2604 wroteRoamingScott wroteTheodoric wroteCan you compare AF performance in AF-C and single point?Caveat: the only similar lenses I have right now are the 85/1.8 on Z and 90/2.8 macro on FE.
I did this AF-C test at 2.8 on both, with both set to highlight meter.
Nikon failed to acquire the background in both dynamic and single point, and had this odd glitchy behavior in dynamic:
Sony paused a bit after losing the initial focus point but quickly found the background every time.
In a broader sense, I use wide mode (area equivalent on Nikon) all the time, with great results on Sony. The camera is sticky and can easily pick out obvious subjects. Nikon's area mode is one step above non-functional compared to Sony. I can see no use case so far where I'd pick area vs a single point.
If you can manage to keep the very small single point box, or slightly bigger dynamic box on the subject on Nikon, it seems to hold its own. I'd like to see more options on Nikon, like Sony's "center" or "zone", or at least be able to change the size of the single point box, like Sony's 3 flexible spot size options.
Single point seems roughly about the same (the lens will play a large role here) on these 2 lenses, with less focus breathing/settling on Nikon. The Nikon seems more sure of itself in AF-S than Sony, which is no surprise...Sony's AF-S mode is a no-go zone for many shooters because of its unreliability.
It really does feel like you're trading one mode for the other between these two bodies.I believe the dynamic focus area extends beyond where the dots end so you likely still had the focus point on the can? That was weird.Check my other vids, does the same thing with single point.
Yup edited an additional response above.
Buckeye2604 wrote: RoamingScott wroteBuckeye2604 wroteRoamingScott wroteTheodoric wroteCan you compare AF performance in AF-C and single point?Caveat: the only similar lenses I have right now are the 85/1.8 on Z and 90/2.8 macro on FE.
I did this AF-C test at 2.8 on both, with both set to highlight meter.
Nikon failed to acquire the background in both dynamic and single point, and had this odd glitchy behavior in dynamic:
Sony paused a bit after losing the initial focus point but quickly found the background every time.
In a broader sense, I use wide mode (area equivalent on Nikon) all the time, with great results on Sony. The camera is sticky and can easily pick out obvious subjects. Nikon's area mode is one step above non-functional compared to Sony. I can see no use case so far where I'd pick area vs a single point.
If you can manage to keep the very small single point box, or slightly bigger dynamic box on the subject on Nikon, it seems to hold its own. I'd like to see more options on Nikon, like Sony's "center" or "zone", or at least be able to change the size of the single point box, like Sony's 3 flexible spot size options.
Single point seems roughly about the same (the lens will play a large role here) on these 2 lenses, with less focus breathing/settling on Nikon. The Nikon seems more sure of itself in AF-S than Sony, which is no surprise...Sony's AF-S mode is a no-go zone for many shooters because of its unreliability.
It really does feel like you're trading one mode for the other between these two bodies.I believe the dynamic focus area extends beyond where the dots end so you likely still had the focus point on the can? That was weird.Check my other vids, does the same thing with single point.Yup edited an additional response above.
I still don't think it's contrast. AF-S has no issues with those same dark zones, it's only AF-C that struggles here.
JadedWriter wrote:
What exactly would you be photographing that you think the MC105 wouldn't be able to handle it?
Skittish wildlife for which every second counts.
I'm not so worried about the lens but about the camera hunting around and being unreliable when I need it to be on point.