Fujifilm X T5: Performance Tests - no ads, no interruptions

A detailed and hands-on performance review of the Fujifilm X-T5, including focus speed, burst tests, high ISOs and more. This video is not sponsored and not interrupted by ads.

UPDATE
**********
Viewers have provided some suggestions to improve Burst/AutoFocus performance:

CONTENTS
**********
00:00 - Intro
00:33 - Apologies
01:27 - This video's settings
01:52 - High ISO (JPEG)
03:14 - High ISO (HEIF)
03:57 - Focus
05:07 - Burst test
06:54 - Burst and AF-C
08:03 - IS for stills
08:33 - Video settings
09:24 - Video quality test
09:58 - Video AutoFocus
12:15 - Recording duration
14:10 - Exposure changes
14:43 - Conclusion

RELATED VIDEOS
**********
**********
I have not received compensation in exchange for posting this review.
If you would like to support me, members of this channel receive additional perks:
Or support me directly via PayPal
While I do appreciate your support, no content will ever be behind a paywall.

FLICKR
**********
The images used in the video, (including EXIF) are available on flickr

ABOUT ME
**********
I do read and respond to all comments, which are moderated. If you leave a relevant question or civil comment without links (which are filtered by youtube) it will be posted, I will reply.
Spend more time with me:
CONTACT: maartendotheilbronatgmaildotcom

DISCLOSURE
**********
I am not sponsored or compensated by any Fujifilm Canada or any manufacturer, I have not accepted payment to review this or any product. Fujifilm Canada loaned me the camera for review, however, I don’t allow them (or anyone) to review the script or video prior to posting.

I am compensated by Google Adwords, who place videos before and after my videos. I receive a small commission from B&H and Amazon if you purchase using the affiliate links.
Please do not allow this to influence your purchase decision. Should you wish to support me, please use the links above.
I encourage you to visit and support your local photo/video retailer.

[ServerSideSquid]

Even though I'm settled on my X-H2, I still thoroughly enjoy watching your videos of other models. You have such a great presenting style which make your videos so much more watchable than other reviewers.

The simple but pleasing 3 note melody as you carefully place down the toy you were looking at so curiously, before addressing the viewer with a friendly catchphrase, is such a good touch.

[michaelwhiles]

Thank you for your review which is detailed and helpful. We’ve all suffered the Ravers and the Haters ! Recently updated to an XT 4 from an XH 1 with no plans for a further move but a ‘ proper’ review such as yours is always welcomed 🇬🇧🇺🇸🇬🇧

[dougmacmillan]

Your tech videos are always detailed but clear. You are honest about likes and you present potential negatives without all the hysterics that are all too prevalent.

[bmeclipse]

Maarten, excellent review, as always. I found your AF-C test interesting. I’ve read others who complain about Fujifilm’s X-T5 and X-H2S autofocusing. I watched another YouTube AF-C test of the X-T5, showing limitations of the AF system. Thanks to the myriad of settings, I’ve used trial and error in my X-T5 and X-H2S, but I have not once seen an affect as the one you experienced.

It could be differences in lens, lighting, or simply firmware.

I have had situations where zone AF focuses on background objects or the occasional subject not be detected, but not an AF that seems to go nap-nap mid shot. I’ve seen other YouTube videos where the X-T5 would not detect a subject. When watching these presentations, I noted a similarity: the AF would lose focus when some sort of subject detection (animal, bird, or Face/Eye) was used… on inanimate objects.

After seeing this, I tested Subject/Eye Detection on my daughter then back to a wooden matryoshka. What I found was that the AF would always pick up my daughter’s head or eye and lock focus. With the matryoshka, it flickered on and off, mostly missing face and or eye and instead focusing on zone detection points.

Using animal detection, I tested this on my dog and my daughter’s stuffed toy dog, which looked very similar in height and color. AF locked onto my dog and tracked him without fail. The stuffed version, not so much. The X-T5 didn’t seem to want to detect its head or eyes, zone focusing on other body parts, but would occasionally detect a head and only once detected an eye.

Right after I watched your video, I tested AF-C, Animal Detection on my dog, XF 33mm f1.4, ES, 20fps, 500ISO, 1/250, f1.4 and at f5.6 moving toward the camera. I had no issues with focusing whatsoever until my Cavie moved to within minimum focus distance (the animal detect was still tracking his eye even within this distance).

I realize that other manufacturers’ AF subject detection might indeed lock onto inanimate objects that mimic what the subject is supposed to be. For AI to work this way, however, should be the exception, not the norm. I would expect Subject Detection AI to work the way it does on my Fujifilm cameras, detecting actual, live subjects.

It reminded me of my time on submarines in the US Navy. The sonar detection computer utilized a catalog of sounds in order to distinguish between marine life, surface vehicles, and subsurface vehicles. It could accurately display the exact ship, its heading, it depth (if applicable), and its speed. These were exact fingerprints of marine craft.

I imagine AI algorithms work in much the same way. There are points of detection that must correlate with detection characteristics within a database. Distance between the eyes, eye movement, cheek bones, eye depth, ear-to-eye distance, eye-to-hairline distance, facial movements, lip depth, lip length, beak size, beak shape, muzzle distance, and on and on.

I went into using the new systems expecting this to be the case and I wasn’t disappointed. I absolutely love having subject detection on my Fujifilm *finally.* But I understand other people’s experience is not my own. I can only judge based on my on experience. I had a learning curve, even though I already used Fujifilm cameras, and getting focus and exposure wasn’t immediately straightforward on the X-H2S as it was with the X-T5. Merely a difference in system and inexperience with it.

I expect Fujifilm will have firmware updates (like for eye detection that gets the eyelashes and not the eye), but so far, I’m enjoying the new experience. Just finished two weeks shooting in London and Paris. I have no complaints.

[DalsPhotography-Daniela]

Dear Marteen, I thank you sooo much for these kind of unbiased videos :) I have always waited for your videos before buying a camera. My best wishes for you and the family, from Uruguay.

[erikurlings]

Erg informatieve videos, heldere info, goed gebracht.

[patrickcanullas]

too good to have your review sir martin. im a fan. 👍🇵🇭 it's my wishlist camera, , xt30 goin'x to xt5. great great review 👍

[NeilMcAliece]

This is the first time I've seen a difference in reviews X-T5 vs X-H2 beyond the spec level. Well done on the detailed review.

I was very surpised to get a 1/2 second shot with the X-H2 in low light with a vintage British motorcycle parked in the middle of the street with a bus passing through behind it and 2 helpfully still people waiting at the bus stop.

This was on a Sigma lens without OIS. I lamented not having a tripod with me, but thought "let's give this a try"

[catherinegrimes]

This is a very interesting video.
It is also the first video that I have seen that tests the HEIF capability of Fujifilm cameras. Judging by your reaction when comparing JPG and HEIF files, you seemed to be a bit puzzled by the perceived image loss in image quality at high ISO and slower burst rate when using HEIF.
The continuous autofocus tracking was surprising as well where the tracking was lost when taking photographs.
Whilst HEIF images are supposed to be better and have smaller files, however, more processing is required to produce these files. Maybe this could be the bottleneck in burst mode? The loss in tracking when taking photographs could also be a symptom of processing speed. Maybe the performance of the processor is slower on the X-T5 than the X-H2? It could be just the software.
The perceived loss in image quality is puzzling. Maybe the tests could be repeated, but this time comparing JPG, HEIF and RAW on the same scene to determine whether the perceived image quality is inversely related to the theoretical image format quality? This problem could be a software error. Maybe your your test scene a high ISO settings confused the algorithm?
The plan is for me to upgrade from my X-E3 to something like the X-S10 or its replacement because of the IBIS, better focussing, more features, the ability to take videos without changing the settings and the large grip that everybody seems to love. If the replacement had HEIF, it seemed to be an attractive proposition with smaller higher quality images. In theory the burst rate would be faster as well. These features would be wonderful for wildlife photography. According to your tests, it is worse in both respects. Maybe Fujifilm have gone a step too far with these 40MP cameras because there are more pixels to process and this is affecting the performance of the cameras?
When the replacement for the X-S10 is released, I will have to be very careful when looking at the reviews and I look forward to viewing yours.
Hopefully, all of these problems will go away with a firmware update.

[fusion-frosty]

Thanks for looking at HEIF and the detail oriented review. I've been keeping tabs on the new Fuji YouTube reviews for a few months now and I believe this is the first comparison with HEIF that I've seen.

Lower HEIF Buffer wise; I'd think the additional processing/RAM usage could be the cause, as others have mentioned. Even with compressed RAW, some other camera model tests have shown shorter bursts or taken longer to clear the buffer after.
Image quality wise; I'd be interested to see a dedicated comparison with RAWs in the mix. The perceived differences could also be down to increased colour depth as well as Fuji's settings/implementation.

This has piqued my interest on HEIF, I will definitely look into the format further. I'm sure there're a lot of general comparisons out there and current adoption rates would be important - with these newer/complex compression schemes it also takes more power to decode, one of the viewers I use took a while to get a plugin out and does seem to take longer to open the few HEIF files I've downloaded.

[aggro_online]

The difference you experienced in AF tracking performance between the X-T5 and X-H2 is concerning and not something I've seen before. You would think they'd at least be within a slim margin of each other. Would you consider doing another controlled comparison of this where both begin at their factory default settings?

[Powerland]

The AF issue perhaps a question of the buffer? No Tracking, when the camera is saving? Buffer is much bigger in the H2. Otherwise the question: Boost ON? Fokuspriority AF-C ON? MF+AF OFF? All these three parameters are essential for a functioning AF in Fuji cameras.

[Roman]

6:15 I'm not an expert but I believe HEIF images use more advanced compression techniques, and require more processing power to decode and display than JPEG images, which could also contribute to a slower camera performance when shooting in HEIF mode.
Some cameras may be better optimized for HEIF and may not experience as significant a performance hit as others.

[_LYH_]

6:08 is it likely the camera processor works harder to further compress the HEIF format, thereby dropping its fps quickly?

[rickslijkhuis]

That AF burst fail at 7:26. Isn't that a firmware issue? Why should the tracking fail on the X-T5 and not the X-H2? I would assume that the buffer limit should not interfere with the tracking but it seems it does.

[LLIMEJIb]

About overheating X-T5.
I can see red sign overheating after 2 hours 40 minutes recording. After next 20 minutes I stop recording manually.
Conditions for testing:
In Settings set to High Temp.
X-T5 USB charging.
Camera stay on wooden table.
About 23-24°C in room.
6.2k30 h.265 422 LongGOP 100mbps (for longer time record).

About quality settings:
Earlier I tested how long X-T5 can record on one fully charged battery.
6.2k h.265 422 LongGOP 100mbps - about 2 hours 10 minutes.
6.2k h.265 422 All-Intra 360mbps - about 2 hours.

[shanwen]

My XE-1, XT-2 and XH-2 (same sensor as XT-5) all the same image quality, at least tested up to ISO 6400. So that's one point to take note. Buying newer Fuji cameras is more for the faster AF and functionality.

[just_eirik]

What about noise in RAW files? How does it hold up? Especially compared to the X-T2 and 3.

[swiftblades]

here's what I'm wondering: is this a worthwhile upgrade over the xt-4? I typically shoot candid photographs in lower light scenes and I sometimes struggle to have enough light in my photographs and the higher megapixel count in an aps-c sensor concerns me slightly that this could be worse relatively speaking.

[scottmarshallphoto]

Love your videos Maarten. Do you have a way to contact Fujifilm to report your AF tracking issue that you experienced when you were shooting images? I feel this should be an easy fix in firmware since the XH2 has the same internals as the XT5.