Monday, January 6, 2014

An unfair comparison? -- the Fuji X-E2 vs the Sony A7r

On the face of it, the idea of comparing the results from a smaller APS-C 16mp sensor to those from a larger full frame 36mp sensor is a completely lopsided affair, and something I did not intentionally set out to do it. Quite by accident, while testing the Sony A7r for another blog review I intend to post later this week, I discovered that some images taken by the lower 24mp Leica M 240 were really quite close in quality to the A7r, and this started me wondering about how much of an improvement a high resolution camera like the A7r would be over something like the Fuji X series. So, just for the fun of it, I set up a comparative situation to see how close I could come to the full frame image quality of the Sony A7r while using the smaller sensor of the Fuji X-E2..


I chose a situation where I could use a reasonably high ISO -- in this case I settled on 1600, although I did the test in a range from 800-6400.  The real problem with comparing two unequal sensor sizes is that the final images are not the same size, but what I wanted to find out was how the images would compare if they were the same size.  This meant I had to either up-res the Fuji image to 36mp or down-res the A7r to 16mp.  I chose to go the high route and make both images 36mp.

My preferred method to up-res an image is to do it in Adobe Bridge where I can utilize the full 16-bit RAW file information in the process. Comparing the two images showed what I expected, namely, that the image from the A7r would naturally show a higher resolution than the one from the X-E2. What surprised me was that it did not show a significant difference in noise between the two at the 1600 ISO.  So my next thought was: Could I perform a gentle massage to the Fuji X-E2 file that would bring it closer to the image quality of the A7r file? For this I went back to Bridge.

This is the naturally room lit scene setup for the test images. It was photographed with both cameras set to ISO 1600 and f/5.6. You can download the Sony A7r file and the up-res 36mp version of the Fuji X-E2 file with the links below.

In Bridge on the X-E2 RAW file I gently adjusted the sharpening and detail tools until I had a reasonable facsimile to the A7r image. Adding the sharpness, small as it was, increased the contrast and started to lose some of the highlight detail. To fix this I lowered both the highlights and the contrast until it looked right.

For the X-E2 RAW image I set the Resize to fit option to 36.2MP in the "Workflow Options" area of Adobe Bridge. This produced an up-ressed version of the Fuji 16mp file to equal the size of the A7r file.

Sharpening sliders in Adobe Bridge were set to 55 for Amount, and 50 for Detail. Luminance Noise set to 16, and Luminance Detail at 50. 
One difference between the two cameras was their lenses. For the Sony I used the Sony 28-70mm  zoom. This is not a very good lens and contributed some softness to the image. Nonetheless, it is the only native Sony lens available for this camera with a proper focal length for this test. On the Fuji X-E2 I used an equivalent 18-55mm zoom, which is far superior optically to the Sony lens.

Both cameras were on a tripod. For the Fuji I used a standard, screw-in cable release. For the Sony I used a 10 second delay timer because the camera only accepts a proprietary release that I didn't have.

Conclusion:

The purpose of this exercise for me was to see if I could approach a 36mp, full frame image quality of a superb camera sensor like that in the Sony A7r using a Fuji X camera and applying some moderate post-processing techniques. I actually surprised myself with the results. I did not go into this experiment expecting to come as close as I did to equaling out the two images.

There is an old adage that says: "It isn't the camera; it's the photographer."  What I think this experiment illustrates is that proper camera technique and good use of post-processing are as important as the camera itself in producing top notch results.

I also found the X-E2 a far easier camera to use. The Sony A7r has an annoying two-step process for changing the focus point, something that Fuji thankfully corrected with firmware awhile ago. The Fuji menus in general are easier to navigate, and niceties like the good old standard screw-in cable release socket make a photographers job that much easier.

The point of this exercise was not to see if one camera was better than another. Instead, I wanted to see if the Fuji X-E2 with its handicap of smaller sensor and lower megapixels could approach the level of a top full frame model. You can judge the results for yourself by downloading the images above. As far as I am concerned, the Fuji X-E2 came through like a champ.

15 comments:

  1. HI Tom,

    I made a comparison of the X-Pro1 and a D800E. I think the A7R sensor is very similar to the D800E. Both are from Sony.

    Do you shoot landscape or stills in the studio with lots of details?
    My experience is, that I only see a difference between both cameras if I shoot with the D800E and very,very good lenses under optimal circumstances with a tripod and so on.
    And, of course, if I do large scale printing (24x36inches).

    All in all Fuji's technology is amazing.
    In my view the only successor over Fuji is the D4 sensor from Nikon.
    But we have to pay around 7000,- €/$ for a D4/Lens combination.


    ReplyDelete
  2. Bought an X-E1 as a travel camera, next to my Canon 5D2 with Zeiss lenses....
    Did the crazy thing and compared the images ...
    Yes, there was a difference... maybe one could say that the Canon was a bit better :-) (that was before I learned about Iridient Developer/Capture One)
    Sold the Canon. Am very happy now. Bought a second Fuji, the X100 (classic, not S). Even more happy.

    That is really weird, isn't it? I am, like you seem to be at leas a bit, someone who has found more joy in photography thanks to Fuji.

    Of course I do believe that the Sony's are awesome camera's, but from every report I have read it seems that taking full advantage of 36 MPixel is quite a task.

    ReplyDelete
  3. Ishan - I use a D800 primarily for landscapes and stills also, and, like you, found that with such a high megapixel sensor comes a big responsibility to supporting it with good technique and first rate optics, and the camera must always be on a tripod.

    This test with the Fuji is going to have me using it a lot more and taking even greater care in my post-processing of images. - t

    ReplyDelete
    Replies
    1. Hi Tom,

      I have a Zeiss 135mm ZF2.
      This lens supports the D800/E very well.

      Most of the time I use manual focussing.
      I think about selling my Fuji lenses and by more Zeiss lenses.
      Actually i consider to by the Zeiss 35mm ZF.2.

      I use a Kipon adapter and shoot with the 135mm and the X-Pro1, so I have ca. 200mm.
      The IQ is very good.
      Zeiss said, there is no problem with APS-C sensors.
      With the 35mm ZF.2 I have ca. 50mm for the X-Pro1.

      Do you use 35mm/85mm lens on your D800?
      And do you have experience with adapters uising with Fuji cameras?

      Ishan

      Delete
  4. Ishan - You might want to check out the blog post I did using Nikon G lenses with a Metabones adapter on the Fuji X cameras: http://aboutphotography-tomgrill.blogspot.com/2013/08/metabones-speed-booster-nikon-g-lens-to.html

    I do use both the 35mm and 85mm on my D800 and D4 - both excellent lenses. - t

    ReplyDelete
  5. ok, 1600 ISO is an affordable value for almost every modern APSC sensors. try the same at 6400 :)

    ReplyDelete
  6. Good point, Nicola. Glad you brought it up.

    Fact is, I did try the experiment at ISO 3200 and 6400 with similar results, but to my way of thinking, anything above 1600 is an impractical ISO range to be using with a 36mp sensor. I use a Nikon D800 regularly and would never use it to go above 1600. It's a bit of a waste, if you think about it, because you are throwing away the best qualities of a high mega-pixel sensor when using it at high ISO's. A better question might be: How would the comparison work at LOWER ISO's, where a 36mp sensor should have a greater advantage. - t

    ReplyDelete
  7. and also you can ask: and what if you use the same upsample trick the Sony result to 80 MPx? :)
    anyway, as a Fuji user who sold my whole Nikon equipment a few months ago, I like the result.

    ReplyDelete
  8. Exactly! I was waiting for someone to figure this one out and post it! Thanks. :)

    Maybe I'll give it a try.

    For me, the most important thing to come out of this little experiment is that I learned something about gently improving smaller image files in post-processing, and having the 36mp image as a target to hit kept me from going overboard.

    ReplyDelete
  9. Not surprised in the least. I returned two A7's (not R) as I was totally underwhelmed by the high ISO performance. At least in jpeg (I rarely shoot RAW, I have my jpeg worflow down pat and it works wonders for me). Of course one cannot state this on the internet and not get immediately taken down in flames, lol...

    ReplyDelete
  10. That Sony is one hideous camera...

    ReplyDelete
  11. This was impressive - and I own the X-E2! I've been tempted by the Sony A7, not because I need full frame, but because of its small size and the Zeiss lenses. The Fuji 35 1.4 is fault-free, but it sort of lacks character, while many Zeiss lenses have a certain something ...

    ReplyDelete
  12. I'm a argentic photographer... When i read the test i'm back in the 80's... Back to real photography!

    ReplyDelete
  13. Tom,

    What an eye-opening comparison. Learned something new. 10/10.

    ReplyDelete
  14. When I look at the exif data. I noticed the shutter speed of the A7 is one stop faster than the xe2. Is this something that we should worry about the fuji? that the ISO meter is not accurate.

    ReplyDelete