What's the deal with the X-T5 autofocus?

yomimoi

Active member
Messages
50
Reaction score
102
OK, first off I know this post is bound to raise suspicions, what with this being my first post here, but please bear with me.

I'm looking to upgrade from X-T2 now that Fuji has finally released the almost ideal camera I'd been waiting for and, although there are a couple other things that bother me, the AF issues (if that's what they are) are definitely stopping me from rushing to the shop. I went through a similar situation with my Pentax K5 a about a decade ago, and I don't want to get burnt twice.

I've seen one, maybe two threads so far discussing Omar Gonzalez's and Jerred Z's findings regarding autofocus, but what worries me is that I keep finding new reviews on YouTube that mention that there are definitely many instances where the AF confirmed focus with the green box, but the photos were subsequently found to be out of focus, or at least not focused on the eyes as confirmed by the camera. See for instance these three:




These are all professional photographers or at least people who look like they know what they're doing. In at least one of the videos above, there are also several different comments from users describing the same discrepancy between focus confirm in camera and the end result.

I know several forum regulars are quite happy with the X-T5 here, and even the above reviewers who've run into problems agree that the AF is a definite improvement over that in the X-T4, so that's good.

However, for event photography, weddings and other similar situations, you just can't afford to have 5-10% (or whatever the exact percentage may be depending on the lens used, AF parameters and shooting conditions) misfocused shots because many of those are unique moments that must be captured well in focus, no matter what.

So I suppose my question is, what do you guys think? Is the AF firmware kind of undercooked and likely to get a big improvement in future firmware updates? Or maybe Fuji should issue detailed guidelines covering specific shooting scenarios and types of lenses, to help users minimize errors? I don't know.
 
OK, first off I know this post is bound to raise suspicions, what with this being my first post here, but please bear with me.

I'm looking to upgrade from X-T2 now that Fuji has finally released the almost ideal camera I'd been waiting for and, although there are a couple other things that bother me, the AF issues (if that's what they are) are definitely stopping me from rushing to the shop. I went through a similar situation with my Pentax K5 a about a decade ago, and I don't want to get burnt twice.

I've seen one, maybe two threads so far discussing Omar Gonzalez's and Jerred Z's findings regarding autofocus, but what worries me is that I keep finding new reviews on YouTube that mention that there are definitely many instances where the AF confirmed focus with the green box, but the photos were subsequently found to be out of focus, or at least not focused on the eyes as confirmed by the camera. See for instance these three:




These are all professional photographers or at least people who look like they know what they're doing. In at least one of the videos above, there are also several different comments from users describing the same discrepancy between focus confirm in camera and the end result.

I know several forum regulars are quite happy with the X-T5 here, and even the above reviewers who've run into problems agree that the AF is a definite improvement over that in the X-T4, so that's good.

However, for event photography, weddings and other similar situations, you just can't afford to have 5-10% (or whatever the exact percentage may be depending on the lens used, AF parameters and shooting conditions) misfocused shots because many of those are unique moments that must be captured well in focus, no matter what.

So I suppose my question is, what do you guys think? Is the AF firmware kind of undercooked and likely to get a big improvement in future firmware updates? Or maybe Fuji should issue detailed guidelines covering specific shooting scenarios and types of lenses, to help users minimize errors? I don't know.
I'm sure with the anecdotal and "real" reporting as well as the other grumbling going on it has definitely reached Fuji's ears. I agree for events, pros cannot live with a high percentage of missed shots. That's why I do NOT use a brand new cam as my primary tool for a paid event. There are too many issues that could go wrong. Admittedly I violated my own policy by using the X-H2 for 30 headshots only 3 days after I received the H2. However I had 2 backups and one could easily determine whether or not the AF was okay.

Fuji has a staunch reputation of dealing with issues via firmware. Hey, they updated the FW in a lens released 10 years ago!

I am not concerned.... but if you are a pro and do events, I would be wary of something brand new. Even the Nikon Z9 had issues that were eventually addressed via FW updates.

--
Bob aka BobsYourUncle
DPR Co-MOD - Fuji X Forum
 
Last edited:
I had the XT3 and XT4 and used them quite intensively.i tested the xt 5 last weekend in order to know what the XT5 did for me.Don’t worry too much , the XT5 is a real upgrade on IQ ánd AF.I ordered mine
 
I think the question you should first ask is "What is the deal with these Youtubers that struggle to take sharp photos when other do?"

I shot with the X-T3, a camera that some of these same photographers struggled with for 3 years. As a nature and wildlife photographer specializing in avian I do a lot of birds in flight and that is a very demanding type of photography. The X-T3 delivered wonderful images for me time and time again and I still use it as a backup to my X-H2s. As I could get 90% or more of my flight photos in focus I don't understand how a camera that these same photographers say is better than my X-T3 can't do at least as well as me on simpler subjects.

There are a few event photographers on this forum and possibly one of them will chime in.

Morris
 
I also upgraded from a T2 (and H1). It is miles ahead of what the T2 could do and overall I'm very happy with it. There are still bugs though (even just navigating the menus) so I imagine it will see at least a handful of improvements. I was always hesitant to use CAF with the T2 but with the T5 it's way better. I'm not sure if the hit rate would be good enough for pro work, but for me it's fine (shooting a toddler and hyperactive dog till now). Really I need to shoot more with it to get used to it since there are so many new AF features that I am not accustomed to.
 
I think the question you should first ask is "What is the deal with these Youtubers that struggle to take sharp photos when other do?"

I shot with the X-T3, a camera that some of these same photographers struggled with for 3 years. As a nature and wildlife photographer specializing in avian I do a lot of birds in flight and that is a very demanding type of photography. The X-T3 delivered wonderful images for me time and time again and I still use it as a backup to my X-H2s. As I could get 90% or more of my flight photos in focus I don't understand how a camera that these same photographers say is better than my X-T3 can't do at least as well as me on simpler subjects.

There are a few event photographers on this forum and possibly one of them will chime in.

Morris
I tend to agree with Morris.

However, I wonder if the AF-C issue is the difference between what one might describe as the traditional single green box approach versus the newer eye/face/animal/etc approach.

Last weekend with my XT5 set to AF-C, focus priority, 7fps, single AF Mode, Custom Setting set to 6 (0, 1, Front) photographing our daughter's dog running towards me every image was sharp (used a 50/f2). Ok, hardily a challenge as the dog is small so his speed was not that fast. I placed the green box on the dog’s face, kept it there and it was fine.

It is worth noting the role of the green box. It is, according to Page 55 of the manual, to confirm that “the camera is able to focus.” That is different from focus being accurate, although there is a reasonable presumption in the mind of the photographer that able to focus and focus accuracy are the same thing (I am sure they are on other brands). This is also the first time I have come across the term false positive, and I can see how intensely frustrating this is.

Is it that these more complex algorithms are not quite there yet in Fuji-land, but which could be with more research, development and firmware updates. If that is the case, they should have been completed before its release otherwise a perception that focus is an issue will be reputationally damaging. The genie is almost out of the bottle and if there is a false positive issue to resolve, then Fuji had better fix it very quickly.
 
At least of the YouTube reviewers mentioned above had the shutter priority in Release rather than Focus ... So it would be expected that some of the photos would be out of focus.

Mario
 
I had the XT3 and XT4 and used them quite intensively.i tested the xt 5 last weekend in order to know what the XT5 did for me.Don’t worry too much , the XT5 is a real upgrade on IQ ánd AF.I ordered mine
It's great to hear it works for you. I think no one disputes it's a real upgrade in terms of AF and resolution (IQ not so much, it seems?). However, if the AF confirmation misbehaves as found by some reviewers, it wouldn't work for me. Let's see what happens as more people get the camera and Fuji irons out the kinks.
 
I agree. It's early days and it's better not to use new gear for critical jobs. Still, Fuji new a vastly improved AF was a much awaited feature for this new generation of cameras, so I wish they'd nail it prior to releasing the camera. Had they done so, everyone would be singing the praises of the improved AF system.

We'll have to wait and see.
 
I don't think it is realistic to expect a 100% hit rate with ANY camera during a wedding or other event. That never happened during the days of film cameras, and of the hundreds of weddings I photographed over the years, I never delivered a "perfect" set of images, yet no complaints. I had used the X-T3 professionally for several years, and even with its "outdated" autofocus capabilities, it still delivered the goods for numerous editorial assignments without issue. Using the X-H2s and X-T4 mainly now for assignments, focus accuracy is just not a major issue, and when I blow a shot, is it is usually user error.
 
I don't think it is realistic to expect a 100% hit rate with ANY camera during a wedding or other event. That never happened during the days of film cameras, and of the hundreds of weddings I photographed over the years, I never delivered a "perfect" set of images, yet no complaints. I had used the X-T3 professionally for several years, and even with its "outdated" autofocus capabilities, it still delivered the goods for numerous editorial assignments without issue. Using the X-H2s and X-T4 mainly now for assignments, focus accuracy is just not a major issue, and when I blow a shot, is it is usually user error.
I agree. However, the issues seem mostly to be associated with the T5 and H2......Sony, Canon, and Nikon, all not 100% perfect, have set a very high bar for continuous AF "in higher MP cams." So it is not an unreasonable expectation for new users of the T5 or H2 to get a higher hit rate than some are reporting. Of course, there will always be user error.

All that said, I am not getting a worse hit rate than (I hate to admit this) than an A1 or Z9. But I have not put my H2 through any truly demanding events.
 
Sorry, no. I respect their work and abilities just as I respect yours (and I mean it). From seeing their work and watching a few of their videos, I doubt they're all as incompetent as you're implying.
You need to acknowledge that many of these Youtubers tested them right out of the box without understanding their machines. And you will often find the 1st comment (or they may have pinned them to the top) is someone reminding them how to set up their camera correctly.

I had my X-T5 for 3 weeks and used it for work in an event, did not notice that false focusing issue. However I have experience unstable WB and exposure shift, and I think the later is more of user error. But for AF I think it is a huge upgrade from my X-T3 that I used for work over 4 years.

However I do reckon Fuji can, and should, make them more "fool proof", work right out of the box without further setting needed........like Sony and Canon.

--
http://www.instagram.com/pennyfan.photography
 
Last edited:
See for instance these three:




These are all professional photographers or at least people who look like they know what they're doing.
or professional you tubers.
 
None of them strike me a being professional. They are not even using a strap, which is a big red flag right there.
 
OK, first off I know this post is bound to raise suspicions, what with this being my first post here, but please bear with me.

I'm looking to upgrade from X-T2 now that Fuji has finally released the almost ideal camera I'd been waiting for and, although there are a couple other things that bother me, the AF issues (if that's what they are) are definitely stopping me from rushing to the shop. I went through a similar situation with my Pentax K5 a about a decade ago, and I don't want to get burnt twice.

I've seen one, maybe two threads so far discussing Omar Gonzalez's and Jerred Z's findings regarding autofocus, but what worries me is that I keep finding new reviews on YouTube that mention that there are definitely many instances where the AF confirmed focus with the green box, but the photos were subsequently found to be out of focus, or at least not focused on the eyes as confirmed by the camera. See for instance these three:




These are all professional photographers or at least people who look like they know what they're doing. In at least one of the videos above, there are also several different comments from users describing the same discrepancy between focus confirm in camera and the end result.

I know several forum regulars are quite happy with the X-T5 here, and even the above reviewers who've run into problems agree that the AF is a definite improvement over that in the X-T4, so that's good.

However, for event photography, weddings and other similar situations, you just can't afford to have 5-10% (or whatever the exact percentage may be depending on the lens used, AF parameters and shooting conditions) misfocused shots because many of those are unique moments that must be captured well in focus, no matter what.

So I suppose my question is, what do you guys think? Is the AF firmware kind of undercooked and likely to get a big improvement in future firmware updates? Or maybe Fuji should issue detailed guidelines covering specific shooting scenarios and types of lenses, to help users minimize errors? I don't know.
I've watched and heard about some these videos too and I own an X-T5 (and before that the X-T4, X-T3, X-H1 and X-Pro 2) and I'm happy with the improved AF speed and accuracy of the X-T5 and haven't experienced anything of the likes yet.

That being said, I rarely use AF-C or Face detection so I would have to do some more testing I guess.

EDIT: just saw some people in the YouTube comments mentioning that this erratic behavior goes away when you turn off AF+MF.

I've never ever had this set to ON so I can't comment on this.
 
Last edited:
Sorry, no. I respect their work and abilities just as I respect yours (and I mean it). From seeing their work and watching a few of their videos, I doubt they're all as incompetent as you're implying.
You need to acknowledge that many of these Youtubers tested them right out of the box without understanding their machines. And you will often find the 1st comment (or they may have pinned them to the top) is someone reminding them how to set up their camera correctly.
I agree and this is a common problem with reviewers. If they don't want to invest the month in learning and setting up a camera them self, they could at least do a google search, find a popular setup and use it and state the setup they are using.
I had my X-T5 for 3 weeks and used it for work in an event, did not notice that false focusing issue. However I have experience unstable WB and exposure shift, and I think the later is more of user error. But for AF I think it is a huge upgrade from my X-T3 that I used for work over 4 years.

However I do reckon Fuji can, and should, make them more "fool proof", work right out of the box without further setting needed........like Sony and Canon.
I agree and have stated that Fuji should make the camera work more out of the box or provide setup guides for different types of photography. The last I know Sony takes some setup to make work well as does Nikon. Canon seems to have it usable out of the box yet there is room for tweaking.

Morris
 
Last edited by a moderator:
Well thanks to those who provided the usual *constructive* comments (“meh, professional YouTubers…they don’t know what they’re doing”…). That’s exactly the kind of feedback that’ll make anyone spend 2,000 € in a camera without thinking twice.

Me, I’m a little more cautious with my money and will wait until more users have reported on their experience, and maybe Fuji addresses this.

I suppose I’m a little puzzled and disappointed with the whole thing. While the AF system is, by all accounts, an improvement on previous ones, it doesn’t seem to be, at least in its current form, the implementation that I expected would end this long-running debate. I’m still hopeful though.
 
There were early online grumblings about AF issues when X-H2/T5 released across online site/forums. Fans quickly swarmed in countering with it works amazing for them etc...

Jerred Z made video showing the AF issues he encountered. He used eye AF af-c with "release priority" (default). Changing to "focus priority" would be better as many comment. However, his response was his sony/canon works perfectly fine with "release priority" and shouldn't need to switch.

Later Omar Gonzalez showed people that AF+MF needs to be turned off as warned in the manual which many would over look.

I'm sure there will be more AF quirks and best settings yet to be discovered. Sit tight, wait, let the early adopter/guineapig figure it out and/or maybe Fuji will address it via firmware before then. If not there is always X-T6/H3 or go Sony/Canon. IMO, this AI AF stuff Fuji added is new tech and plenty of improvements will be made.

My hunch is the processor in the X-T5/H2 isn't quite fast enough to handle some AF situation to perfection at 40MP. AF-S is fine, but AF-C had some issues. X-H2s seems fine since it is only 26MP? Might be one reason why Fuji decided to refresh the EVF of the AF box position less. Maybe algorithm is inefficient or needs tweaking. Shrugs. Sony BIONZ processor was already speedy handling eye AF better than prior Fuji. When they added AI AF chip to A7RV and the processor got upgraded to be 8x faster.
 
Last edited:
I think the question you should first ask is "What is the deal with these Youtubers that struggle to take sharp photos when other do?"
Again, no, you’re misrepresenting facts. They don’t “struggle” to take sharp photos. If you watch the videos you’ll see plenty of sharp photos (at least as far as a YouTube video allows to assess).

Their issue is with the miss rate, which seems to be noticeably higher than with other systems they have used and tested, and the fact that the green box seems to be giving too many false focus confirmations. I think it’s a problem worth looking into. If it’s not a problem for you, that’s great, but I like to be a little more respectful with the experience, abilty and needs of others.

Anyway, I’ve read enough posts in this forum to know there are, like in any brand-centric forum there ever was, blind apologists on duty. So I won’t take this any further.
 

Keyboard shortcuts

Back
Top