Fuji X-trans sensor too sharp?

Messages
3,209
Edit My Images
Yes
Comparing side by side Canon 5D mark 2 against Fuji X-E2 shots I took on Saturday at airshow, I'm seeing tiny blurs on a lot of my Fuji photos while Canon photos are mostly okay.

The Fuji images are really nice and sharp when photographing stationary objects. But nature of fast jet is that they are moving really fast. I found even with 1/500s I am still getting that ever so slight blur, not as sharp as stationary objects, not as sharp as I know Fuji could produce.

With the Canon, I found the images look similar no matter stationary or moving. So the camera feels more forgiving, at same shutter speeds to Fuji, I can get shots I'm happy with, shots I feel it is sharp enough, shots I would happy to crop heavily for web use. (eg down to 3megapixel)

Canon 5D mark 2 with 70-200 F4 IS, IS mode 2.
Fujifilm X-E2 with 55-200 F3.5-4.8 OIS, OIS turned on, does it auto detect panning? should this be turned off for panning with fast jets?​

Are Fuji X-trans sensor images too sharp, thus similar to Sigma DP2, it is unforgiving for wrong techniques?



On a similar note, I have found with similar photos (X-E2 with 10-24mm at 23mm F8 landscape shots), I much prefer Fuji X100 (also at F8) photos compared to X-E2. The X100 images look very pleasing while the X-trans sensor look almost too sharp, details become distracting from photo composition.

So are X-trans sensor images just too sharp?




P.S., I'm using Lightroom 6, if that makes any difference. I heard LR isn't great at X-trans RAW files, but that was when X100s was released, should be okay now?
 
I have to admit to not being impressed with Fuji raws when I used Adobe Camera Raw however I noticed a huge difference in how the raw files came out using some of the other options, PhotoNinja was very good but if I had a Mac (soon) I hear the Irident Developer is meant to be the best option.
 
In LR 6.1 apparently they have fixed some and are planning to fix other in future versions http://www.fujirumors.com/adobe-say...-edge-definition-ps-cc-lr-cc-and-lr6-updated/

We have reduced “color blur” artifacts when processing Fujifilm X-Trans raw images. In collaboration with Fujifilm, we are still investigating methods to improve fine detail rendering and overall edge definition.”

Sounds like made images look even sharper!

Sorry, I am already using Lightroom 6.1. Not 6.0.

So it's purely a software rendering issue?
 
I always had issues with my X100S and how it worked with Lightroom. Now it's at 6x it seems far better but still not on the same pace as most of the other Raw files I have (Sony, Nikon, Leica, Oly). The lens on the X100S wasn't outresolving the sensor so i couldn't comment on sharpness besides it appeared similar if a little duller than a Leica X2, but not as sharp as an EM-1 with a decent lens. Still plenty fine enough for everything really.
 
Processing may well have an effect but I think technique does too, especially if switching between full DSLR and tiny Fujis. I found I got blurry pictures to start with from my Fuji because the smaller camera wasn't as easy to hold steady but that seemed to go away as I got more used to using it.
 
Have to say that I would not consider fast AF to be a particular strength of Fuji cameras. I have not tried-out XT-1 but my XE1/2 bodies were not fantastically good in AF-C mode when used for wildlife shooting.

Maybe this is a contributing factor ?
 
If you want to see whether it's lightroom's fault, then you could always use the in camera raw converter and compare shots.
 
Anyone know if the 55-200 lens OIS supports panning? Should I have turned it off?

I'll upload a few photos, it's definitely motion blur, F11 at near infinity should not have any near-miss AF problems.
 
Anyone know if the 55-200 lens OIS supports panning? Should I have turned it off?

There is no panning support in any Fuji OIS systems, you need to turn it off if panning.
 
Anyone know if the 55-200 lens OIS supports panning? Should I have turned it off?

I'll upload a few photos, it's definitely motion blur, F11 at near infinity should not have any near-miss AF problems.

Ok, I obviously missed the f11 at near-infinity element. That stated, I still do not think that AF acquisition is particularly fast, in relative terms to a decent DSLR, on the XE2.
 
Here are the photos, both are actually taken very similar time, I'm using the X-E2 and my wife (novice who only took it up because I had not sold the Canon gear yet) using the 5D2.
The Fuji shots are ever so slightly blurred, Canon looks sharper.

100% Lightroom view comparison:
ImkGBK4.jpg


Canon


Fuji
 
It could just be that your wife is a better photographer than you :) I have much better camera gear than my wife but she often gets better results than me :(

Did you shoot raw only or have you also got in-camera jpegs, and if so, are they also blurry in the same way? That would rule out the LR question.
 
It could just be that your wife is a better photographer than you :) I have much better camera gear than my wife but she often gets better results than me :(

Did you shoot raw only or have you also got in-camera jpegs, and if so, are they also blurry in the same way? That would rule out the LR question.

Noooo, that can't be right :p

In-camera JPEG are also ever so slightly blurred.

I heavily suspect I shouldn't have left OIS turned on while panning for these jet aircrafts. I wrongly assumed due to lack of IS mode select, the Fuji 55-200 is able to detect panning like the Canon 24-105.
 
There is no panning support in any Fuji OIS systems, you need to turn it off if panning.

You learn something new every day! I really should reread the manuals...
 
Can I ask why you're using 'f/11 near infinity' (or f/8 with the larger format?).

Why not simply use AI Servo on the Canon and the Fuji equivalent, open the aperture up a bit (its only the centre of the frame that need to be sharply in focus anyway) and therefore, increase the shutter speed? I've found even fast jets at airshows are quite easy to track as they mostly simpley go left to right and vice versa.

It seems you're sacrificing shutter speed for no reason!

Neither shot looks as good as I would expect them to!
 
Focusing isn't Fuji's strong suit, so I thought F11 at infinity should mean it would ensure sharp photos most of the time. The Canon was on P mode since my wife doesn't do numbers. Canon was using AI Servo, Fuji was using single shot because continuous focusing means continuously hunting for focus.

No, neither are good photos, there are better ones. According to time stamp, both are taken at similar time, so I used it as comparison.



Panning isn't really the main issue I don't think, WW1 plane photos also have very similar problem. Many of the photos that day is like this, there are tiny blurs when zoomed in.

This problem is pretty consistent. It's also seen with 60mm macro lens when shooting flowers that move in the wind, if the flower isn't perfectly still, there's always some slight motion blur seen even at 1/200s. I think it's because the sensor is very intolerable to tiniest movements when the photo is taken, probably due to the pixel layout and lack of AA filter.

I've heard Sigma DP2 Merrill have similar problem, really really high resolving sensor = not very user friendly.
 
No comparison to other cameras this time, but you can clearly see the 1:1 version is not as sharp as it could be.

Both are click-able for original size I uploaded, the original is straight out of camera JPEG.

1:1


original
 
Perhaps the lens is a bit iffy? Or diffraction is taking hold at F11. Presumably the planes are far enough away that you can just infinity focus to take the AF guesswork away. There are a few forums on DPReview with people complaining about the lens, But then again, people do love to complain rather than endorse! :-D http://www.dpreview.com/forums/post/50984479
 
F/11 shouldn't cause diffraction on the 5d2 but f/8 may on the Fuji?
 
The sharpest point on the Chop is cut off in the enlarged portion. It is somewhere near the edge of the chop which is the middle of the frame.
 
Diffraction may be the cause of soft images, but I still think the sensor is just too sharp leaving no room for slightest hand-held problem.

Take a look at the 1:1 image below by clicking into Imgur. 1:1 is ever so slightly not sharp due to the slight camera shake. I have to take a few to get one that is sharp. Sensor too sharp?


No comparison to other cameras this time, but you can clearly see the 1:1 version is not as sharp as it could be.

Both are click-able for original size I uploaded, the original is straight out of camera JPEG.

1:1


original
 
Diffraction may be the cause of soft images, but I still think the sensor is just too sharp leaving no room for slightest hand-held problem.

Take a look at the 1:1 image below by clicking into Imgur. 1:1 is ever so slightly not sharp due to the slight camera shake. I have to take a few to get one that is sharp. Sensor too sharp?


Poor logic there some where......
Camera shake and the like are on top of the image that the lens produces. Both an sharp image or a soft one are subject to the same amount of camera shake, or subject movement. The extent of the blur produced by such movement will be the same for both images. However the sharper the images was to start with the sharper it will remain.

If you use the new shake reduction filter in Photoshop You will be able to see if the problem is caused by shake, as it is able to make a massive difference to the sharpness. Though at the cost of perhaps some white line artefacts.

I suspect that you are indeed suffering from some camera shake in these pictures, which may well be reduced if you shoot as a burst rather than a single shot.
 
Last edited:
All looks like a tiny amount of motion blur to me. Have you tired comparing something with edges using a decent shutter speed and OIS on/off? The 55-200 is a very good lens but will be past it's best at the long end and and past about f9, plus add in a teensy bit of instability and voila..soft soft soft.

This is a 100% crop from the OOC JPEG - same sensor and lens, 200mm f5.6 1/350 OIS on.

View attachment 44310
 
Back
Top