Fujifilm X-T1 Field Test Part III
Fuji X-T1 Review -- Field Test Part III
Of hot rods and blimps...
By Mike Tomkins | Posted: 06/17/2014
During my two weeks in Hong Kong, I had plenty of opportunity to test out the Fuji X-T1 in all manner of shooting situations -- or at least, most of them. There wasn't much opportunity for shooting sports or active subjects, so I saved that for my return to Knoxville, TN.
And then my PC died, leaving me scrambling to repair it, and causing me to miss my planned sports shoot -- but fortunately there was a car show in town once I was done with the hassle of resurrecting Windows.
To test the X-T1's autofocus chops, I took it to a car show. The gorgeous retro body looked right at home in the company of vintage cars like this 1969 Chevrolet Camaro RS/SS 396!
The Hot Rod Power Tour in Knoxville's Chilhowee Park promised lots of color and visual interest, but at first blush, not a lot of action. As it happened, though, I found a good spot from which to shoot near the exit to the car park as all the hot rods, classics, and not-quite-so-classics (including everything right down to the infamous Chevy Nova) were leaving for the day.
I settled in at end of a short straight on the outside corner of a turn, so that the cars were coming straight towards me, and shot with the 55-200mm lens to frame tightly.
From the time the cars took to traverse the length of the road, I'd estimate that they were moving at 25-30mph, which while not spectacularly fast, still meant a fair amount of focus adjustment to track with each vehicle. And as an added bonus, the speed wasn't constant either -- there was first acceleration towards me while I was trying to achieve a focus lock, followed by deceleration for the corner. (And I panned to follow each car through the start of the turn.)
Straight away, I hit two problems. One I should have expected -- that frustratingly easy-to-bump exposure compensation dial. The other I wasn't expecting -- a significant viewfinder lag that made framing difficult. This could've been a great shot, but I cut the car off just slightly due to the lag.
Straight away, I hit my old "friend", the Fuji X-T1's easily-bumped exposure compensation dial. Reviewing my first few shots, I accidentally dialed in +2/3EV exp. comp. -- not enough to immediately notice through the viewfinder, but enough to clip the highlights under dappled sunlight through trees, since I was shooting JPEG-only at the time.
Fortunately, I caught the problem after a few cars, but it was a bit of a shame to lose these images, especially since one of the prettiest cars at the show was amongst them.
Another unwelcome surprise in those first few shots was the X-T1's viewfinder lag. I've praised it in my earlier shooter's reports, and for ordinary shooting it's exceptionally fast-reacting, by electronic viewfinder standards. When shooting bursts in Continuous High drive mode, though, it's much more bothersome.
I think what's actually happening is that the X-T1 isn't showing the live image between frames at all, just a review of the last image shot. It's something we've seen and commented on with other cameras before, but I must admit given the X-T1's otherwise-swift viewfinder, I wasn't expecting to see it here.
Although the viewfinder lag during burst shooting means it wouldn't be the ideal sports camera, the Fuji X-T1 did a reasonably good job of keeping up with these cars head-on at around 30mph.
To try and quantify the scale of the problem, I used another camera to film a video showing both the X-T1's live view feed and a flashlight positioned so that it was visible to both cameras. Then I stepped through the video (shot at 60 frames per second) and compared when the camera filming the video first showed light from the flashlight, and when it appeared in the X-T1's live view.
When simply framing with the X-T1 but not shooting, there was precisely one frame of 60p video between the flashlight switching on or off, and the live view feed showing it happen. During Continuous High shooting, though, the delay soared to a full 10 frames of 60p video, or around 1/6th of a second.
That lag made it quite tricky to frame accurately with the X-T1 in burst shooting. What my eye was seeing lagged what my hands were doing -- and with a fairly powerful telephoto, even the slightest twitch of my hands made a significant change to the scene. The result, initially, was that I'd overshoot my intended framing adjustment, then once I realized I'd done so a sixth of a second later, I'd overshoot a correction in the opposite direction. I felt more like I was chasing my subjects than framing them.
It reminded me of another situation with a Fujifilm connection, from a decade and change ago. At a trade show in Las Vegas, NV, Fuji were kind enough to invite me on their blimp, and (this being a pre-9/11 world) the pilot invited me to take the controls for a little while -- with close supervision, of course. Shooting with the Fuji X-T1 reminded me of flying that blimp, and of the pilot's advice on correcting my initially-rather-erratic course.
After a little while, I got used to the lag and learned to think ahead of the camera. With that experience under my belt, I could frame tightly as I wanted, but it did take some practice.
Blimps take a while to respond to control movements, and so in my brief flight time, I had to learn to overcome the urge to respond to what I was feeling happen, and think ahead of the craft. I had to do much the same with the X-T1 -- think ahead, and fight the urge to respond to what I was seeing in the viewfinder.
While I'd have been happier zooming and panning to follow these moving cars with a DSLR that showed the real scene between frames, it was possible to learn to work with the lag, and to get at least reasonably close to the framing I was going for. That is, until the cars were well into the corner, just 30 feet or so in front of me, and I had to pan sideways very quickly. At that point all bets were off, but by then it was also too tough a challenge for the autofocus system most of the time anyway.
Which brings me nicely back to autofocus -- the very thing I'd originally been intending to test. And how did the X-T1 perform? I'd say the results were fair, but not stunning. Perhaps 20-30% of my shots didn't quite nail focus, but the overwhelming majority were at least close. (And I was pixel peeping here -- printed or viewed full-screen, I'd say nine out of ten images were sharp enough.)
Also, when focus tracking did drift a little astray, it usually returned to a good focus within just a frame or two. Only once in hundreds of frames did it run wild, going extremely far out of focus. That was when I was panning to follow the car I was shooting, which was almost at its closest point, and moving quite quickly relative to the camera. Excluding that one instance, though, focus stayed at least close to the ballpark.
The classics kept rolling and I kept shooting, scoring some really nice results straight out of camera. Fuji's JPEG rendering was just to my taste, more so than Adobe Lightroom and Camera Raw's defaults.
Another thing I particularly wanted to get a handle on before rendering my verdict on the X-T1 was raw file processing. This is the first X-Trans sensor-based camera which I've fully reviewed, but I'm aware that for quite some time, third-party support for Fuji's unusual X-Trans color filter array was rather less than optimal.
My understanding is that the reasons for this were twofold: one being that supporting X-Trans in the first place required a pretty significant development effort that would only pay off for a relatively small number of photographers, compared to the market for Bayer-filtered cameras. That, coupled with the complexity of demosaicing the X-Trans array -- and the fact that many image processing algorithms used by third-party apps had been developed with Bayer cameras in mind -- meant a lengthy delay for support in the first place, and some image quality issues even once support arrived.
It was my understanding that this had been largely resolved as X-Trans has matured, but really, I wanted to see for myself. As I'm predominantly a Lightroom user, I started there, but I also looked at Photoshop, DxO Optics Pro, and researched the situation with some other key apps.
Using Adobe Photoshop Lightroom 5.4, images from the Fuji X-T1 imported without any fuss at all, and I didn't notice anything untoward in Adobe's rendering of the images. Obviously, Lightroom does render a little differently from the out-of-camera JPEGs. I found that the in-camera profiles -- Provia, Velvia, Astia, Pro. Neg. and Monochrome -- which were recently added by Adobe make things quite a bit closer to their out-of-camera equivalents, though. (And these really are nice to have, if you're a fan of the camera rendering.)
There's something about classic American muscle cars that really gets my pulse racing in a way that modern cars just don't. This 1965 Pontiac GTO is a very nice example of the breed.
Sadly, these aren't available in Photoshop unless you're a Creative Cloud subscriber. I've so-far resisted upgrading because I haven't seen any feature compelling enough to persuade me to pay an ongoing subscription, so I'm still on Adobe Photoshop CS5. That means I can't open the X-T1's .RAF raw files out of the box -- first they have to be converted with Adobe DNG Converter.
I used version 8.4 to do so, then had a play with the settings, but immediately noticed the absence of any camera profiles except Adobe Standard. Photoshop Camera Raw's rendering of the images was otherwise much like that of Lightroom, which isn't really surprising given that they're based on the same underlying engine.
I had a feeling that DxO Optics Pro wasn't going to touch the X-T1's raw files, as I knew last I'd checked that the French software company hadn't adopted support for X-Trans. Sure enough, that's still the case, and there's no sign of it changing any time soon.
My understanding is that the reason for this is DxO's unique algorithms which so beautifully correct for a wide variety of defects, but are closely tied to the presence of a Bayer CFA. Reworking these for the X-Trans CFA would need a major rewrite, and DxO hasn't yet seen sufficient demand to justify all that work. Nor can Optics Pro handle converted DNGs from the X-T1.
And then... this appeared. I have no idea what this hot rod's lineage is, but it's certainly eyecatching!
The great shame here, though, is that means X-T1 owners won't have access to DxO's excellent PRIME denoising engine, which in my opinion is the best in town at the moment, albeit a bit slow because it works by throwing vast amounts of processor power at the problem.
I was pleased to find that the other main players -- Apple's Aperture and iPhoto, and Phase One's Capture One -- all support the X-T1 in their current releases. I'm personally not a fan of Capture One's interface, and I don't own any Mac hardware, so I didn't give these a spin. It's my understanding that, like Adobe, these competitors also do a fairly good job at this point.
So, it seems that with the sole, sad exception of DxO Optics Pro, third-party support is now up to snuff. That's good news, and it leaves me with only one other feature I wanted to return for another look at: Wi-Fi support. I'd mentioned in a previous Field Test that I used Wi-Fi quite a bit on my trip, but this was with the more basic of Fuji's two available apps, "Fujifilm Photo Receiver".
The Photo Receiver app simply allows transfer of images from camera to phone, with your selections made on the camera. A second app, "Fujifilm Camera Remote", is more comprehensive, providing not only for image download, but also for remote browsing of the camera, geotagging of images, and remote control.
By day's end I got fairly good at framing despite the lag, zooming out and panning to follow cars into the corner with the tight framing I wanted. Here, I believe, an early 70s Pontiac Firebird Formula.
Although I wasn't a big fan of the app's interface, I did quite like the remote control capability. Range was fairly good -- perhaps 50-60 feet with a wooden floor in between camera and phone seemed about the limit before the live view feed became unusable. (But it reestablished seamlessly, as soon as I moved back a bit closer to the camera.) And lag was relatively short, as well -- perhaps a little more than a tenth of a second. The ability to tap on the live view feed to select a focus point is also nice.
The layout is a bit bizarre, though. There's only a very small window in which to see the live view feed in portrait orientation, and the app won't rotate to landscape orientation. Most of the screen around the live view feed is packed with controls and status displays. I'd have expected tapping on the status displays to provide the option to change exposure variables, but it doesn't.
Instead, a single exposure variable is highlighted beneath the status display -- either aperture or exposure compensation. Left and right arrows select a value, and clicking on the button between lets you select the other variable. Beneath is a Shutter button control, with two buttons on either side. The left button pulls up a shooting menu, and the right button an ISO menu. Yet the very top item for adjustment in the shooting menu is ISO sensitivity, even though it has its own dedicated button.
The shooting menu provides access to options such as film simulation, white balance, macro, flash mode, and self-timer -- but not shutter speed. And it seems there's no way to access this function at all, nor to choose to shoot in Program mode, so you're limited only to Aperture-priority shooting. (And this is with the camera body itself set with all dials in their Auto positions.)
No awards for guessing what car this is! A beautiful mid-60s Ford Mustang in the late afternoon sun...
There's also a slider which, when touch-dragged, switches between still and movie capture, and a playback button to take you to the image browser. This last is a nice touch, because it allows immediate transfer without having to disconnect from the camera and re-pair. You can select multiple images at a time for transfer, but sadly there's no way to see anything larger than a tiny thumbnail. (And my HTC One X+ phone has a relatively large screen; many phones would should the thumbnails at an even smaller size.)
The Receive option on the main screen of the app provides much the same interface as the standalone Photo Receiver app, so there's really no need to have both apps installed. Browse Camera, meanwhile, provides basically the same interface used for selecting images from the Remote Control mode, again with only tiny thumbnails.
And finally, the Geotagging function piggybacks on your phone's GPS receiver and uses this to determine the capture location of each image. After tapping the geotagging button in the app's main menu, the phone's GPS receiver is enabled and location logged for up to 60 minutes. This I couldn't get to work, even though geotagging was enabled in the camera's menu.
Even if I could, though, I wouldn't really find it useful. My phone already has short-enough battery life without running its GPS receiver for an hour at a time just in case I happen to take a photo -- and I really don't feel like waiting for Wi-Fi to connect so I can start and stop the GPS service ad-infinitum. Geotagging is a feature that belongs in an in-camera GPS receiver or a dedicated GPS accessory, not in my phone.
But be that as it may, the more expansive Wi-Fi app did prove worthwhile, if only for the remote control capability and its mirror of the Photo Receiver app features. I think I'd probably stick with this version simply because, knowing my luck, the one time I uninstalled it I'd find myself out of cellular connectivity and wishing I had the remote control mode available.
And that about does it for my final Fuji X-T1 Field Test. Looking back over my time with the camera, I found plenty to love, and a few things not to like quite so much. Probably key among these are the too-easily-bumped exposure compensation dial, and the hard-to-press rear-panel buttons.
I really did like the size benefit over my DSLR, though, and image quality was pretty good too. And with the aforementioned control quirks aside, there's just something about the X-T1's aesthetic and build that screams "enthusiast camera". The absence of a mode dial -- and presence of auto positions on the other dials -- is incredibly intuitive, and I'm going to miss it when I return to my daily shooter. Warts and all, I found myself reaching for the X-T1 over my other cameras a lot of the time when I went out shooting, and wanted interchangeable lenses.
Earlier in this review, I compared it to another retro-styled camera, the Nikon Df. I think what Fuji has crafted here is an altogether better thought-out shooter than that camera, one in which the retro aesthetic works, rather than gets in the way. Were I in the market for a new interchangeable-lens body -- and willing to switch mounts -- today, the Fuji X-T1 would definitely be on my shortlist.