DPReview TV: The good and bad of Fujifilm cameras

Yes, use Youtube's captions...

@IdM photography: Yes, if only for the occasional unintended hilarity when the captions mangle stuff. I was watching a video about Eritrean cyclist Biniam Girmay and the YT captions mangled his name into “Vietnam Gourmet”.🤣

(This video had “hassle flats” in it).

Name a good one. All camera maker's phone apps are awful.

Stratman: the average app reviewer disagrees. Nikon Snapbridge has a 4.1 star rating in the Play store, versus 1.6 stars for the Fuji Camera Remote.

Canon Camera Connect: 2.3 stars, so very poor.
Sony Imaging Edge: 2.2 stars.
OM Image Share 3.5 stars, so somewhat decent.
Both Panasonic Lumix Sync and Panasonic Image App: 3.0 stars.
Leica Fotos App: 2.1 stars.
Pentax Image Sync: 2.2 stars.

So yeah, apparently most camera apps apart from Nikon, OMDS and perhaps Panasonic are sucky, but Fujifilm's app clearly takes the crap crown.

In my personal experience, the Fuji app was always unstable in connecting (often requiring three attempts to establish connection), and over the past month it's been unable to connect entirely. The vast majority of 1 star reviews complain about difficulty to connect.

DPR / DPRTV: given how many people live their digital lives through their phone, a camera app comparison would be a great service to the camera buying public.

John Bean: Chris replied to my similar comment on Youtube, he basically said the same as you: most of the apps suck, so they don't bother discussing it.

I use the Fuji, Sony, and Nikon apps, and while I don't think the Fuji app is are terrible as a lot of folks seem to think, IMO Nikon's SnapBridge is the best of the lot. I don't have any issues with either. Sony, that's a different story. In my experience, the initial setup on the Fuji app is the trickiest part, and if you don't follow their instructions to the letter, it will never work properly. I had to start over a few times, but once I got it right, it has always worked well for me.

Yes, it’s a terrible app.

"Name a good one. All camera maker's phone apps are awful."

Shockingly?!? The best app I've ever used is the Leica app. Does what it claims it can do smoothly, with some nice little extras thrown in. Fuji, as reported is weak, with Sony being completely useless...

Comparing manufacturers apps could be a great DPReview article, or test.

Amen to that. Nine out of ten times, I can't even get the blasted app to connect to the camera. Kill it with fire, Fujifilm. Start again.

There are solid rumors that Fujifilm is working in a completely new app. When it will be released nobody knows.

I don't find the current app that bad. Mine always works fine. My greatest criticism is that you can't see a decent definition preview of your images before importing them to your phone. If they solved that, I'd be perfectly fine with the current app

Let's talk about the idea of connecting camera with smartphone.

The smartphone as remote display with touch controls to adjust settings in camera, that is the best idea ever. I have used it since 2013 and it has been better than any side articulating screen ever. Even for macro, architecture etc photography it is better.
It works best for video, but as well for stills.

But how come that we can't have modern quick connectivity?

Bluetooth all-the-time readiness to connect wifi for live view, otherwise just Bluetooth for settings changes... So no long time pairing or anything, just tap menu/button for "connect" and get them talking. Smartphone would open correct app automatically, camera would go proper mode.

NFC to help as well.
Camera in playback of video or photo? Touch smartphone too camera and automatic transfer to smartphone.
Multiple photos and videos selected in camera? Copy all to smartphone!
Camera in live view? Go straight to live view in smartphone with controls!

@IdM photography
That's your usage, which I have no opinion. But my work require high power output in bright sunlight in larger area. HSS simply doesn't work in those cases.

They probably don;t think the market is big enough for them. Most Fuji users (like me) aren;t interested in huge tele's as they are too big and heavy. Remember, we got into mirrorless because the cameras are smaller and lighter. Finally, it's a new system compared to Canikon and if you start out with the idea of shooting nature you'd have gone for one of those. Also, the bodies obviously aren't designed for holding big lenses. Horses for courses.

I've noticed that too. Maybe it just takes a while to run into situations where the AF fails. People tend to test in too ideal a situation in official reviews, where AF from 5 years ago still worked fine.

To be fair though, DPReview was pretty critical of Fuji video AF in their new cameras.

With me it's more single precision that matters for AF as I take landscape and street photos , I am getting missed focus with the x-t3 in single shots about 10% of my photos even though I have the latest AF motor in the 18mm f1.4, I hope the X-T5 AF will nail it for precision
As for AF speed, for me, it's more than enough

marc petzold it looks like you had a bunch of first and second generation bodies and one kit lens. Fujifilm might have lost money on you, as traditionally bodies are sold at or near cost and money is made on lenses. I understand being strapped for cash, so I can recommend the 55-200 as it is incredibly sharp for its price point and also doubles as a portrait lens in a pinch. Given how good the 18-55 is up to 45mm or so, the only prime you need in that range is the 50 F2 (very affordable for how good it is). That leaves the ultrawide range open for either another zoom or a prime like 12 or 14mm. The cheap solution is an XC 15-45 just for the wide range from 15 to 18. You can now probably get rid of 5 out of 7 of your third party lenses.

Yes I figured you're on a budget. Otherwise there's always the 10-24 or Samyang 12mm. I use an adapted Canon 10-18 for my ultra wide needs.

You literally listed that you own only the kit lens plus a bunch of manual focus lenses. My fault for taking your word for it, I guess. Not sure how that elicited such an outburst.

Yes, indeed.

The problem with MF systems is how limited lenses options are.

Lenses are overall very good, but if you don’t like the rendering of a lens covering a specific focal length you are in trouble. Which means that lenses always have a very neutral rendering designed not to displease anyone. Which means they are a bit boring and not very ambitious compared to the best FF lenses by Nikon.

At least with the Fuji you can adapt DSLR lenses. Some work great such as Otus 100mm f1.4.

About adapting, how's vignette affecting the image? Is it acceptable?

It depends on the lens and on the aperture, the Otus 100 is very good on 90% of the frame I would say.

You are not alone...

I do wish that the new cameras are a commercial success
so that we finally get a super hybrid next-gen camera one
more year down the road.

I even started thinking that: maybe the X-T50 or X-E5 may
come to the rescue of videography shooters, but the crop
factors already implemented on the X-T5 -to differentiate
the X-H2(s) models as top of the line- are not good omen.

No need to defend yourself, you have not been under attack :-)
I repeat: I do believe you, no issue. I am just totally confused why some, like you, see these effects and others, like me, don’t see it at all. Just for reference: X-Pro1, x-Pro2, X-T1, X-T3, X-H2s user here with a bun h of zooms and primes between 10 and 200mm

I am thinking hand holdable. But I never had any issues with hand holding my GFX100s at 1/125. I will take some leaf shutter lenses if available.

@biggercountry - In a video like this the faces of the people are the primary focal and exposure areas. The background is unimportant.
For even less distracting background, they could have done it indoors with a white background.

No you are wrong. XTrans got no advantage, if so, at all, why did they skipped it with their "medium format" cameras? Proof taken...

I checked the results and the RAWs were good, but it was a new sensor back then and the same with bayer would be equally good I think. The drawback? Years of no support in Major Tools like Lightroom/DXO which are my tools of choice ;(

they skipped it in MF because of the significantly higher resolution…according to Fujifilm. As I said, moire is much lower and as such, anyone claiming no benefit and just marketing is wrong.

It is only marketing. Cut the same resolution from the center of the medium format cam and compare it. It's a shot in the foot to them and they didnt make the same mistake twice (kudos for that). You need something unique on the market but the XTrans was nothing worth considering outside of the fuji-forum. I had one to test it out and love Fuji since my GX680.

When I say you are wrong…you are wrong.

https://www.dpreview.com/galleries/6343025702/photos/4327089/04c457d5-3fc4-4dde-84a4-9be78ccd528c

https://www.dpreview.com/galleries/6343025702/photos/4327088/f500b6b5-fcd0-4b31-8b85-6b6c53ef5828

If you were correct, the moire between these two cameras would be identical. Because the moire is not identical, you are wrong. End of discussion.

The biggest benefit of X-Trans is that you can get the same (or similar) performance as a Bayer but without the need for an AA Filter, and can mean slightly sharper images since you don't have an AA filter, from my understanding of it. however, some cameras like the Nikon Z50/Zfc and a few others can get away with a Bayer sensor and no AA filter, but I think it's just they might be more prone to moire than ones that have an AA Filter (but it is still possible to get moire from a camera with an AA filter too).

It may have other benefits, but it has its place. Most companies have just adopted the Bayer array as it's most common. Of course the caveat of the X-trans is not all RAW processing software can (properly) deomasic image from X-trans sensors or it leads to some side-effects in the demoasic'd file.

While the X-Trans sensors might reduce moire, they definitely introduce bizarre "worms" and other out-of-focus induced strangeness. Some of those effects might be post-processor related. But it is still a frustrating limitation.

No, they don’t introduce worms…that is poor demosaicing software. Long debunked.

I can easily find the squiggly worm artifacts in many of my Fuji captured images rendered in Adobe Lightroom. I would not call Lightroom "poor demosaicing software" since it does very well with images from other manufacturers. Definitely not "debunked" as you say.

Unless those files were shot and edited by you, it doesn’t prove anything. There are tonnes of guides on the internet for how to properly edit RAF files in Lightroom and not see artefacts, if people choose to ignore them and keep hammering that sharpen slider that’s on them, not Fuji.

I think the worm thing is a bit of both. I can reproduce worms from my Nikon files in Adobe's products if I wanted to. While I'm sure the X-Trans doesn't help, part of it (I feel) is also some RAW processors don't demosaic properly. People have gotten different results with different RAW processors, which may suggest it's not just the sensor.

I think a good example is if you went back to say 5 years ago with Adobe Lightroom classic and compared it to now, wit ha Fuji RAW file, you might see a difference. Unfortunately, I do have some Fuji RAW's from an XT2 and from about 2018, but i don't have an older version of LRC to compare it to with the latest version I have installed. But perhaps someone who has both could do a comparison. I (and many others) strongly feel it falls more on the RAW processor than the sensor though. I suppose one could run some RAWs through the most current version of C1 and LRC and compare.

@MediumFormatLover, while the lower colour resolution of x-trans CFA means its filter pattern is less likely to interfere with fine patterns that are being captured (hence less moire) compared to Bayer, it would also mean higher colour noise level at the same ISO (It has been deliberately made hard for us to compare due to Fuji's persistence to continue using the different ISO standard from everyone else).

The demosaicing algorithms for x-trans also have to be a lot more complex, good ones are so computationally expensive that Fuji's own image processing chip inside the latest X-H series can't even handle them and it rather uses primitive algorithms that introduce fairly noticeable worms in SOOC images.

Some say x-trans produces more random, film-like noise at high ISO that is more pleasant to look at - which I also disagree because the noise pattern looks almost identical between x-trans and Bayer when I process files both CFAs using a cutting-edge RAW demosaicer such as DeepPRIME.

the x-trans gives no benefit to IQ, in fact, the opposite ..but it is getting better with newer camares.. to cut a long story short i had an XE-2 and the results were tearable so i did a very controlled test with the same scene on a tripod with the same sharp prime on the fuji and a 18MP SL1/100D..i dont think anybody on the net had done a test x-trans vs Bayer using the same lens on both.... i asked the fuji community on DPR to edit the RAW file in different software which they did some were better than my effort some not,at the end i revealed i had a control image from the canon and when posted it CLEARLY had more fine detail, the thread got promptly deleted (happy to say that doesn't happen on the fuji forum today) so i did it again but posted on someone else thread and this 1 was never deleted this time and is still live today. 1 guy came very close to matching the canon, twice, CONTINUE

The mental gymnastics after MediumFormatLover posted test images for all to see are hilarious.

I especially like people who start strong with "you are wrong" and use word like "proof" but then one paragraph down come out with the real reason, a little whine about "my tools of choice" lololol

Reality: darktable is free, feature complete, and has handled xtrans correctly since all the way back in 2014.
If you're too mentally limited to learn a new tool, that's on you.

@DrewRick, MediumFormatLover deliberately ignored some important contexts to argue their points. For instance, DPReview suspects that there is some chroma noise reduction baked into the file, according to its review on X-T4 - essentially nullifying the effectiveness of x-trans at reducing the moire (since chroma NR does suppress moire). More scientific comparison done by PetaPixel (https://petapixel.com/2017/03/03/x-trans-vs-bayer-sensors-fantastic-claims-test/) also suggests that there isn't a life-changing difference between both CFAs in terms of the amount of moire you get.

It is possible to get fantastic-looking, worm-free images out of x-trans (though obviously not using the Fuji camera's internal RAW processing). But the question is: Is it really worth having worse compatibility with photo-editing software to (potentially) get rid of a tiny amount of moire that you might or might not notice in real life, when Bayer sensors are coming to solve this issue by having more MP?

@dave8 - The samples from DPReview clearly show a reduction in moire. As such, your ascertain that there is no benefit in image quality is laughable at best, and simply ignorant at worst.

My working hypothesis is they didn't use xtrans in the GFX 50 because they were too stingy to pay for a CPU with enough power to demosaic 50 MP of xtrans in real time back in 2016 when the 50S was produced. Now they're stuck with the illogical marketing statements they made at the time.

Incidentally, Pentax still includes it's AA filter simulator feature in their latest cameras, Canon recently tried to patent the same idea (lol), and high fashion GFX photographers frequently struggle with moire. Maybe they all know why they're doing what they're doing. Of course that's only with the best lenses in studio conditions with fine garments, us mere mortals with consumer grade lenses shooting wide open for the bokeh don't have to concern ourselves with that.

@vscd - As the samples show I am right…you remain wrong…just as I said

I find it so comical that people get emotionally attached to a camera because they own or want it and think it's the best camera available and defend it like their life depends on it. I've been a pro photographer for over 17 years and I've never used a x-trans sensor for professional work and yet I never had an issue with moire. How did I survive all this time? Come on people don't be so dense. X-trans is solving a problem that doesn't exist. Being different is cool and X-trans is different so it's cool. In 2022 it does nothing but lower fine image quality and rob people of fine detail in their photos when compared to equivalent sensor resolutions. Not to mention having to completely revolve your editing software around your camera. 100% of Professional/high end cameras (including Fuji's medium format) don't use X-trans sensors. Wanna know what the I've never heard from a follow photographer in 17 years? I wish I had a x-trans sensor because this moire is totally ruining my images.

I photographed weddings for years. Moire shows in fabrics with Canon and Nikon gear…I never had the issue with Fuji. The samples show that moire can rear its head and the XTrans helps. As such, it solves a problem that DOES exist.

@vscd - I’m not sure what you are talking about. My first response here has 39 likes so far. If your rebuttal to anyone you don’t like is to claim fakes accounts, you’re already on the losing end of a debate. Well done, you!

@ExtraCredit...... it is a common myth that the one being the first has a huge advantage. Those who enter a market last sometimes have the advantage because they can learn from someone else's mistakes, and they can enter when the technology gets cheaper and better. The inventors and innovators are rarely the market leaders, because they ALSO need to be great at manufacturing and marketing.

This is why late entries like Sony in MILC cameras (2013), Toyota in cars (1936), and Apple in smartphones (2007) are the market leaders rather than the innovators Panasonic (2008), Peugot (1882) and IBM (1994).

My digital Leica M8 is pretty beautiful and it was made in 2005. But yes, Fuji makes nice-looking cameras for sure!

I think I'd vote for Epson in 2004 on that one...
http://camera-wiki.org/wiki/Epson_R-D1
The needle gauges...
(Used to have one BTW.)

Both the Leica M8 and fuji cameras "Make nice-looking cameras". But, the questions is, can they both make just as nice looking images as the best Sony, Nikon, or Canon cameras?
Some people think better looking cameras are better than cameras that take better images. I don't feel the same way.

They are tools and basically all are capable of great results in the right hands.

Jon555, I had the R-D1 and R-D1s and still regret selling them!

I suggest you ask in the Fuji forum; its probably settings issue FYI it will display your max selected ISO until you half-press the shutter speed..

that original 16Mpx sensor was magic.

We have choices - so you don’t have to look into Fujifilm :-)
Just as a side note: if you don’t look into it how do you know it’s bad?

Because it destroys existing workflows and sometimes customers have hopes that vendors change. hey, even Sonycams now have something like ergonomic bodies...

It does destroy workflows? Can we rephrase this please? It disrupts workflows, that does not need to be a good thing nor a bad thing - it’s just a disruption.
You lose something, you gain something. At the end of the day progression needs disruption.
Note - I don’t try to say it’s necessarily the right way, time will tell and for one it will be ok, for another one it will be, well, wormy.
And the great thing is: we as consumers / users have choices. No need to complain:-)