Pages

Wednesday, 4 March 2015

Fuji X-T1 Review Part 1 Image Quality and Performance


X-T1 with 23mm f1.4 lens


In February 2014  I posted an opinion piece on this blog titled ‘Fuji X-T1, an ergonomiccontrarian view’.    

This was based on ergonomic analysis of many cameras and a discussion of the relative merits of the traditional [aperture ring+shutter speed dial+exposure compensation dial] control layout versus the modern [mode dial+control dial] layout.   With respect to the X-T1 the discussion was conceptual in nature.

I recently had the opportunity to borrow an  X-T1 from a family member and use it for a week.  It was fitted with the 23mm f1.4 lens and was running Firmware 3.0.

This is the Camera Ergonomics blog so I will concentrate on ergonomic issues in the next post.

However I will make some observations in this post about image quality and performance.

Both the photos below were made on a tripod with 2 second timer delay.

FZ1000


FZ1000 Crop. Compare this to the X-T1 crop below. Foliage and other details are rendered more clearly without smearing or artefacts. 


X-T1  


X-T1 Crop. I think you can see smearing of details in the foliage together with a JPG artefact like appearance in the pink flowers which was in the original RAW file, it is not a product of saving to JPG for publication. It is not dependent on the lens or aperture used.


Image quality  This has been extensively and favourably reviewed elsewhere, however I did note some issues which I think worthy of note.

Overall the camera does a good job with particularly low luminance noise levels at high ISO sensitivity settings.  It is well suited to indoor use without flash, which is a good thing because the camera lacks a built in flash.

ISO Range  luminance noise.  I tested the X-T1 against my usual camera, the Panasonic FZ1000.
For my testing the X-T1 was fitted with the quite large but optically very nice 23mm f1.4 lens which gave good results right from f1.4. Stopped down a little, images were sharp across the frame.

At low ISO sensitivity setting (200)  the X-T1 showed no noise at all. The FZ1000 at ISO 100 had just detectable noise.

At high ISO sensitivity settings the X-T1 easily outperformed the FZ1000 with a two EV step luminance noise advantage at ISO 6400.   Unfortunately the X-T1 does not offer RAW capture at ISO settings higher than 6400.

Files from the X-T1 at ISO 6400 had about the same amount of luminance noise as those from the FZ1000 at ISO 1600.

I also noted clear differences between the two cameras in the character of the luminance grain.

The FZ1000 produces sharp, clearly defined grains, reminiscent of black and white film of yester year.  Grain from the X-T1 is larger and softer in character, producing images with a less grainy appearance but also slightly less apparent sharpness.

FZ1000 Chart center

X-T1 Chart center.  Lower resolution than FZ1000 with JPG like appearance in fine detail of RAW files


Resolution/sharpness (low ISO sensitivity)  

I tested this by photographing subjects with fine foliage outdoors, several subjects indoors and also my standard test chart which consists of pages of classified newspaper advertisements on a flat board.

For the test photos I had the camera on a tripod, used 2 second timer delay and directly controlled aperture,  shutter speed or sensitivity depending on the requirements of each test run. I did runs with autofocus, manual focus, mechanical shutter and E-shutter.  I used RAW capture converted in Adobe Photoshop Camera Raw.

I got some unexpected and puzzling results, some of which I cannot explain.

* On the test chart  at the same aperture (f3.1) and equivalent focal length  the FZ1000 out resolved the X-T1, even at the edges and corners.

On the chart I have several strips of woven cloth. The FZ1000 resolved the weave detail clearly but the X-T1 did not.

Considering the FZ1000 has a 16x superzoom lens and the X-T1 had one of Fuji’s best primes mounted, that was a surprise. The FZ1000 has more pixels of course, which helps, but in the past I have tested 16Mpx micro four thirds camera/lens combinations which produced better sharpness/resolution than the Fuji.

Outdoors in general photography the X-T1 delivered sharper edges and corners. I can’t explain this unless the 23mm f1.4 has a problem with curvature of the focal plane. But that is not a convincing notion either because in several frames of the test chart the Fuji images were sharper in the corners than the edges, presumably due to optical correction of the curved focal plane.

* In the fine foliage outdoors and in the center of the test chart the X-T1 files exhibit an odd appearance of fine details. They are RAW files but they look like oversharpened JPGs, with little double lines at fine detail edges.  Strangely the edges of the test chart did not show this effect.

* In the outdoor photos with the X-T1 some areas of green foliage had smeared or smudged details while the same areas from the FZ1000 held detail.  This issue has been reported elsewhere and is  I believe, a well known problem with Fuji X-Trans files converted in Photoshop (or Lightroom).

* I have 9 pages of text on my test chart. In several frames I found that just the bottom row of the pages showed marked unsharpness with double imaging effect. With other mirrorless interchangeable lens cameras (MILC)  this doubling may be a sign of shutter shock.  But in my experience the doubling caused by shutter shock affects the whole frame, as you would expect.

I have no idea what might cause the phenomenon to be confined to just one sector of the image. It was most marked at  shutter speeds from 1/10 to  ¼ second and was not evident when I used the E-shutter.

Neither camera nor lens has an image stabiliser.

I did not have the opportunity to pursue this matter further.

FZ1000 sharp lower section

X-T1 blurring with doubling of lower section at 1/10 sec, cause not determined by me.


All I can say is that

1. There are some issues, some of which have been reported elsewhere, with Fuji X-Trans .RAF files converted  by Photoshop Camera RAW and presumably Lightroom which uses the same process.

2. I saw mysterious local unsharpness in test chart frames the cause of which I know not.  I don’t even have a hypothesis about it.

Autofocus Performance  It seems that the arrival of every new camera these days is accompanied by claims that is has ‘the fastest autofocus in the world’. That was Fuji’s claim when the X-T1 was released.

Presumably some of the zoom lenses focus more quickly but with the  23mm f1.4 lens mounted, auto focus speed  is quite pedestrian and slower than other cameras which I have tested recently.

In addition the AF action is accompanied by a fair bit of whirring and clunking in the mechanism.

By comparison the FZ1000 focusses so fast and quiet one is hardly aware that any focus action has taken place.   In most conditions the focus box confirms ‘in focus’ instantly with half press of the shutter button.

I found (single) autofocus in the X-T1 to be mostly accurate however I did notice a few ‘just out of focus’ frames indoors and even outdoors at f5.6. 

I did not have a suitable lens with which to test AF-Continuous performance.

Manual focus   This works very well with the 23mm f1.4 lens on the X-T1. The manual focus ring on the lens is smooth and precise.

Fuji’s implementation of peaking is the best I have yet seen on any camera and  with the Focus Assist button fast, accurate manual focus is possible. 

The family member from whom I borrowed the camera has it set up for manual focus as a first preference for photographing children and other family members.  It works well.

However to note an ergonomic issue I would point out that holding the camera, rotating the focus ring, looking in the viewfinder (or at the monitor) then pressing the Focus Assist button while keeping the subject in frame, requires a bit of juggling.

The problem is that to get onto the Focus Assist button the right thumb must drop away from the thumb support. This might not sound like any kind of big deal to read about but in practice the action of lowering the thumb causes complete release of the hold which the right hand had on the camera.

This in turn requires the left hand to support virtually all the mass of the camera and lens while continuing to rotate the focus ring accurately.

This is not impossible but there are other camera designs which allow the user to do the job more smoothly, with better control and without having to juggle the mass of the camera from one hand to the other.

Shot to shot times  With ‘Image Disp’ in the ‘Screen Setup’ tab in the Setup Menu  set to OFF  (it took me three days to find that, by the way) RAW capture and refocussing with AF on every frame, the X-T1 with 23mm f1.4 shot 10 frames in 7 seconds, giving a shot to shot time of 0.7 seconds.

The FZ1000 made 0.3 seconds shot to shot time in the same conditions.

Summary 

Image quality is generally in line with other current cameras having a 28mm diagonal (APS-C) sensor and is particularly good at high ISO sensitivity settings.

Resolution and detail from RAW files converted with Adobe Photoshop Camera Raw is not outstanding. I read elsewhere that other Raw converters do a better job.

Performance is also generally in line with the camera’s peers.

Given this good but not remarkable performance what can account for the numerous rave reviews which the camera has enjoyed ?

Maybe Andy Westlake’s comments in the Digital Photography Review ‘Shooter’s Experience’ report of April 2014 sum up the X-T1’s appeal:  
There's no doubt that the X-T1 is a camera that will make many photographers drool. With dials and switches to operate almost every conceivable setting, it's almost the antithesis of the typical modern press-button-spin-dial interface (which arguably finds its apogee in the X-T1's most direct competitor - the Olympus OM-D E-M1). Match it up with one of Fujifilm's truly excellent primes like the XF 23mm F1.4R and you get an exceptional image making tool.

I hear two things here:

1. The dials and switches are to drool over. Lucky the thing is weather sealed.

2. It is the antithesis to the modern camera user interface. For every technology advance, there is a counter-movement. Some people just lurrve their vinyl records, or say they do.

In the next post I will discuss whether those dials and switches are useful for something more photographic than resisting  drool.

Post Script:  As I was writing this Digital Photography Review published on 28 February,  an interview with Mr Toshihisa Iida, Senior Manager, Sales and Marketing for Fujifilm.

I thought some of his responses to questions from DPR staff were unusually candid. 

Mr Iida said that consumer feedback revealed three positive things about the X-T1  (then listed four) namely lenses, color, portability and viewfinder.  Nothing here about image quality, performance or ergonomics which are the three cardinal qualities by which I evaluate a camera.

He also said that Fuji wanted to improve ‘…operation, in terms of button layout and so on, autofocus performance and movie image quality’.

I did not evaluate movie performance but I did find that various aspects of operation and autofocus performance could be improved.  Read more about ‘operation’ in the next post.

The next post:  is about ergonomics of the X-T1.















No comments:

Post a Comment