Fujifilm X-M5: “Whatever Fujifilm Did, Autofocus is as Good as on my Sony” – Dangerous Overpromise?
A couple of weeks have now passed since the announcement of the Fujifilm X-M5.
And so far the big controversy has not yet received a final answer: how good is the autofocus on the Fujifilm X-M5?
Now a new Fujifilm X-M5 hands-on review has been published by PavSZ (see above).
In his video, he says the X-M5 autofocus is much better than the one on the Fujifilm X-S20.
And when asked about it in the comments to his video, he says multiple times that he shoots also Sony cameras, and that the X-M5 is just as good as Sony in terms of autofocus.
The problem?
PavSZ does not provide any video autofocus demonstration to support his claim. All we have are his words. So here they are as shared in the comments to his video.
First, I did ask PavSZ about it in the comments if he refers to AF-C or only AF-S. PavSZ says:
both [tested AF-S and AF-C] and yes AF-C is very good. Very fast and sticky.
It uses the learning/predictive algorithm. You can see it working, it detects a person or animal in an instant, puts a box around them and split second after starts looking for eyes. When it finds them, it tracks very well. I was very surprised with it
Of course many called him out in the comments to the video, saying his claim is ridiculous and misleading. This is PavSZ’s answer:
I just love that nobody here has actually used this camera but everyone has got very strong opinion that the AF does not work 😂
I test and use regularly Sony, Nikon, Canon, Fuji and Lumix latest cameras and instantly I can tell that this is way above average. Whatever they did, it works
In reply to another comment he says:
in my opinion this is better than my a7IV
And when asked about a difference between X-S20 besides build quality, he says
auto focus. Insanely good
For now all we have are his words based on his experience with the X-M5, which he could handle for 24 hours.
But hey, we trust people who never touched the X-M5 when they say the AF is trash, so maybe we can trust also those, who actually handled it ;).
And while 24 hours are not a whole lot of time, I am sure if I were to give any of you readers an X-M5 for 24 hours, you’d probably be able to make up your mind rather quickly about its AF performance.
PavSZ Advantage over Other Reviewers
The only potential advantage I see is that PavSZ is testing the X-M5 just before its release, likely using the final production firmware—unlike those who reviewed it in September with earlier versions. So the AF on his unit might be better than the one on all X-M5 tests we have seen so far.
Even so, my stance is, “I’ll believe it when I see it.”
A Dangerous Overpromise?
PavSZ’s claims would need visual evidence; otherwise, they risk creating unrealistic expectations.
My X-T5 had great autofocus pre-firmware—but was it on par with Sony? No, but it was still excellent.
I’d be satisfied if Fujifilm’s AF performance simply matched what the 5th-gen platform offered at launch.
Predicting Fujifilm will surpass this and even rival Sony sets high expectations.
If the new firmware “only” returns AF to its original strong level, people may still feel let down if they anticipated Sony-level performance.
In short: there could be a huge improvement that would make us happy, but because of PavSZ statement some might feel disappointed anyway.
More Recent Videos
Down below I will share a few video tidbits from a Chinese X-M5 review where we see the autofocus in action.
I will embed it twice, as depending on which video you click, you’ll start the video right at different points of the autofocus test. Or access them directly at youtube at this point and at this point.