22
Jan
2013

New X-E1 Reviews (and comparison with the 5D MarkIII)

Image credit: Martin Doppelbauer

Photographer Martin Doppelbauer decided to make a somewhat unbalanced comparison: The Fuji X-E1 (price & specs) vs Canon’s EOS 5D Mark III (price & specs). Sounds unfair? Well, given the amazing performance of the X-E1 there are good reasons to be curious. He shot in RAW, using Capture One (Version 7.0.2) to develop the files of both cameras.

The review is not just a comparison between Canon’s full-frame champ and the X-E1, it is also a review that highlights the strengths and the weak points of the X-E1. Trying to shot at the same ISO settings showed some strange behaviour of the X-E1:

The true ISO value is considerably lower than the displayed value. I have performed some tests in comparison with the Canon EOS 5D Mark III, whose metering works particularly accurate according to dpreview.com. When the X-E1 is set to the same aperture and ISO values, the camera determines a much longer exposure time than the Canon. The extension factor was in average 1.75 (with variations from 1.62 to 2.0), which is equal to three quarter exposure stops (EV).
There are reproducible differences between the various ISO levels. The lower values from ISO 200 to 1600 are too weak by about two third exposure levels (EV) on average. The two highest values ​​(ISO 3200 and 6400), however, deviate by one full exposure value (EV).
This means for example, that the X-E1 works with a real ISO 125 when set to ISO 200 and a real ISO 3200 when set to ISO 6400. This will provide for very good noise performance results in comparative tests in magazines or websites. In reality, however, the noise performance of the X-E1 is actually good but not as phenomenal as it seems. The Fuji always has to select almost double the ISO value compared to properly tuned cameras for a given scene, aperture and shutter speed.

So, what about the comparison? Martin writes:

The images of the X-E1 are of such a high quality that a comparison with the full-frame EOS 5D Mark III seemed reasonable. Both cameras were tested together with their “kit zoom lenses”, the 18-55mm f/2.8-4 on the X-E1 and the 24-105 L f/4 on the Canon.

Image on top: Courstesy of Martin Doppelbauer

Some information about the setting of his test:

For fair comparison the settings of both camera systems should be largely identical. This affects focal length, depth of field and exposure (ISO and shutter speed). Due to the different sensor sizes and Fuji’s exaggerated ISO numbers the matter is not so easy.

Focal length and depth of field (aperture) is converted to the crop factor, i.e. with 1.5. For example, a focal length of 23.3 mm on the X-E1 corresponds to the popular 35 mm on a full frame sensor. An aperture of f/5.6 on the Fuji gives a similar depth of field as f/8 at the Canon. I have always reduced the ISO values by 2/3rd steps on the EOS 5D Mark III.

All images were shot in RAW format and developed with Capture One 7.0.2. In some of the X-E1′s pictures the white balance was adjusted according to the EOS 5D, which I generally found slightly more accurate. All other parameters of the RAW software were left at their default values​​, which is particularly important when comparing noise performance.

The X-E1 can’t (obviously) hold up to the Canon EOS 5D Mark III. Never the less,it is interesting to see how well the Fuji performs, and how little is missing to get an almost full-frame like performance. Quoting Martin’s conclusion:

Regarding resolution: It is to be noted that the EOS 5D Mark III (with its low-pass filter) records visibly more details than the X-E1 (without the filter), even though the pixel count of the Canon in horizontal and vertical axis is just higher by 18%. Obviously the omission of the alias filter does not help the X-E1 to increase resolution much.

[...]

The X-E1 is a camera with impressive mechanical and optical quality and great usability. Its images come close in quality to the EOS 5D Mark III over a wide range of ISO settings.
It is pleasing to hold the camera and taking pictures is great fun. Operation of the X-E1 comes close to classical range finder cameras. The Fuji is perhaps not for casual shooters, but photographers who deliberately compose their images will have great pleasure. For them the slightly slow autofocus will not mean much. After all, it regularly nails sharpness right to the point.

I gladly confess that I was never annoyed by moiré artifacts in real shooting situations. If present at all they were rare and weak. [Update 2013-01-19: I was out shooting in the snow today. Snow-covered trees in bright sunlight are good for great pictures. But I also got a pretty significant amount of colored artifacts.

So, I let it up to you to check pics and to compare them. There are a lot (really a lot) of sample images in Martin's post, as well as test shots of the Siemens star for better understanding of moiré and aliasing artifacts.

Some more X-E1 tidbits:

Fuji X-E1 price check: Amazon, Adorama, DigitalRev, J&R, B&H, eBay

Canon EOS 5D Mark III price check: Amazon DE | Amazon FR | Amazon UK | Amazon US | B&H | Adorama

[via Martin Doppelbauer]

Follow Me
  • http://photos.remektek.com.au Paul Szilard

    I have done a loosely similar comparison between my X-PRO1 and my Nikon D4. Not surprisingly, the D4 beats the XP1, however the size/weight and enjoyment of use, plus unique features still make the Fuji a joy to use.

  • http://andreathode.de andrea thode

    Nice comparison. Interesting thing with the ISO sensitivity! The end result is somehow what I expected. I’ll pass on this generation and think that I’ll get in once the X-Pro2 is out.

    Until then, I need to carry my heavy D800 with me …

  • Markus

    I am not sure how good the “standard” raw setting is as a starting point for comparison. I find it more realistic to enhance details as well as possible without artifacts as a proper comparison point… then again the lenses remain as a difference.

    • stan

      I observed also the wrong iso indication of XE1. In my comparison with samsung nx1000, xe1 indicates 2/3Ev higher iso than nx1000 for the same shutter/aperture ( e.g iso5000 of xe1 vs. iso3200 of nx1000 ). Interesting is that the incorrect iso indication has also olympus OM-D it indicates cca +1EV higher iso as should be. This was confirmed also by DxO Mark test. Seems Fuji marketing tries to sell X-trans sensor cameras as comparable to full frame, and also Olympus OM-D 4/3-rd sensor as comparable to APS-C sensor. Seems that this is not technically possible, then the marketing use such manipulation of ISO to sell better. The clear conclusion is that LARGER SENSOR IS REALLY BETTER, regardless of marketing tricks/manipulations. Nevetheless I discovered this trick of XE1 at my tests, the first reaction was I will return it back ( I still had the possibility ). Afterwards I made more comparisons, and I find out that also considering real ISO ( eg. iso3200 insteed of indicated iso5000-6400 ) the XE1 (Xtans sensor) has still best detail/noise of all APS-C cameras. Now I’m still happy owner of XE1.

      • Markus

        I guess I noticed it myself since I was surprised about needing longer shutter times than expected. That said it’s not all bad for landscapes having a lower ISO though having it marked right is much better.

        My comment on standard settings I was alluding to resolution though I think the kit may he he bottleneck here.

      • http://www.riccimarco.com marco

        same here. a comparison of my XP1 with 5D2, in a controlled environment (studio) show me the XP1 is .5/.66 EX less sensitive then canon 5D2.

        by the way XP-1 as a far (FAR) better DR in highlight and in shadow lift test than my 5D2

        highlights http://www.kaotic.it/trash/TESTDR/
        and shadows http://www.kaotic.it/trash/TEST_DR_SHADOW/

        • Christian

          Thanks for these impressive examples!

          Yeah, I sold my 5DII a while ago because of the limited shadow DR and kept my old 5D…the X-E1 is better than both.

        • http://andreathode.de andrea thode

          Thanks for the examples, really impressive for such a camera.

      • Rune

        You are wrong. 35mm dinosaurs are dying. As for the E M5, here is what dp review had to say:

        By our tests, the E-M5′s measured sensitivities are about 1/3 stop lower than indicated across the ISO range (i.e. images are fractionally darker than expected for any given set of exposure values). A discrepancy this small has little practical impact in real world use.

  • Peter

    The Fuji shots are also consistently wider. So either the Fuji has actually something like a 1.4 crop instead of 1.5 or the reported focal lengths from either the Fuji or the Canon are not the actual focal lengths.

    • Markus

      I think you are confusing the crop with a full image. Since the Fuji has less mpx the same resolution crop is wider.

      • Peter

        Ah yes, of course. Thanks!

  • dave

    A little bit of a disappointment. Fuji Xtrans technology seems to be excellent nevertheless, but there is also a relevant impact of marketing strategy close to misleading customers if the ISO comparison is true. So let me wait for the FF to come – somewhen.

  • http://www.peterwerner.net Peter

    That finding about ISO sensitivity is quite interesting. However, we should not jump to conclusions here. All still photo lenses are rated in f-stops. An f-stop is a theoretical value based on focal length and diameter of the aperture opening. It does not factor in how much light is lost travelling through the glass of the lens.

    It might well be that the Canon lens simply lets through more light at the same f-stop setting. The only way to do a meaningful comparison would be if the lenses were known to be set to the same t-stop (which is also the value a light meter gives you and why still photographers have to “calibrate” their meters to each lens/camera combination). One way to do this could be to put the same lens onto both the 5D and the X-E1 (using an adapter with no glass elements).

    • Renato S.

      I would like to see that, but I think it’s not the first report I read about Fuji cheating in the ISO department and I just got to know that the Fuji uses the same sensor as the D7000, NEX-5n, etc. Although the different color-array can improve sharpness against the same sensor with bayer-filter, it does not make miracles. I had hoped that it was really a new sensor, but it seems that it’s not. Anyway, let’s see what they can bring with the next generation of X-cameras. It’s said that Sony has a new sensor and it will come very soon in the A58, this can be interesting.

      • Adam

        Renato — When photos of some of the X-Pro1 prototypes surfaced on the web, some clever users noticed that the pinout pattern was the same as the Sony 16mp sensor you mention. That started the rumor that Fuji was using that sensor. But when the production X-Pro1s came out, the pinout pattern had changed. The source of the X-Trans sensor is still unknown, but its very unlikely it’s the Sony model.

        • Renato S.

          I’ll dig in a little bit more in this subject but I’d prefer it’s really not a Sony sensor, if not, there is not much of a difference between a bunch of cameras.

          I’d be happy if the next generation of X-cameras could bring – this time – a accurate ISO performance.

  • Scott

    Boy thats some nit picken, But still thats a FF its having a shoot out with! And isn’t far off the mark.

    • http://www.riccimarco.com marco

      that’s true.
      anyway my XP-1 +35 1.4 is about half stop less sensitive then 5D2+50 1.4

  • Gaz

    I was out photographing in the snow with my X-E1 and used a hand held Sekonic light meter for the exposure. The X-E1 was constantly giving nearly the same reading or differed by a 1/3 stop!. Maybe his 5D mkIII is the one with the problem?
    As for the resolution…”The camera doesn’t make a bit of difference. All of them can record what you are seeing. But, you have to SEE”. —Ernst Haas

    I would rather carry my X-E1 around than a full frame DSLR :)

    • cognoscenti

      Or maybe your lightmeter is faulty?

      You need to get both cameras and expose it on the same scene. Then compare the shots. Look at the histograms.

  • Christian

    Very interesting – and the Fuji is doing very well here.

    The ISO sensitivity mismarking was also mentioned in the dpreview.com review of the X-Pro 1, Page 17: “By our tests, the X-Pro 1′s measured sensitivities are about 1/3 to 1/2 stop lower than marked, which is unusual for a modern camera. This means that for any given light level, shutter speed and aperture the X-Pro1 has to use a higher ISO to get an image of the same brightness as an accurately-rated camera.”

    They continue: “The X-Pro1 does an exceptional job of balancing noise and detail in its JPEG processing, even taking its generous ISO ratings into account.”
    And RAW: “That said, the output from the X-Pro1 is remarkably good at high ISOs, and it’s difficult to see APS-C Bayer sensor cameras quite matching it at ISO 6400 even with substantial noise reduction applied. In fact the X-Pro1 is giving the full frame Canon EOS 5D Mark III a decent run for its money here. Needless to say, this is very impressive stuff.”

    I´m happy with my X-E1 ;)

  • Troll Prozac

    I’ve owned a lot of full frame cameras:
    5D
    5D II
    D3s
    D800
    D800e
    1Dx (my current cam).

    Overall the D800e has the best pure image quality. The 1Dx is the best all round “workhorse” camera (which is why I have a DSLR at all, they’re work horses).

    However I use my Xpro1 more than any DSLR I’ve ever owned. Why? Because it’s fun, the colours and contrast of the images are amazing and it is compact enough to fit in my Billingham Hadley Pro :P

    So it’s not just about the mesurbating differences (if that bothered me I would have kept the D800e). In the end these minor differences really don’t make that much of a difference. It’s getting the shot that counts (which is why I love the speedy reliability of the 1Dx for work) and also having fun while doing it (which is why I love the xpro1 for everything else.

  • jon burtoft

    Hi, as far i am concerned the iso thing is a storm in a tea cup. I previously had a panny g1/ gf2 and a canon 40d, my xpro1 is far far better at high iso than those, at iso 6400 compared to the 3200 of the 40d the difference is incredible.
    To be getting such amazing quality files at these iso’s is remarkable.

    In the digital age we are truly spoilt by the advances in technology and what it can offer us. I enjoy using the xpro1 immensely and am more than satisfied with the image quality. It is an amazing testement to the fuji cameras that they are running comparisons with the full frame 5d mkIII.

    There are different tools for different jobs, you just have to pick the right tool for your requirements. Image quality on all recent cameras is fantastic, the biggest difference to me is form factor and usability ie sensor size, external control dials, ovf, size, weight, lenses etc.

    Cheers Jon

  • Ricci

    This ‘Comparison’ is actually one of the best ‘Review of XE1′ that I read so far. It addressed many limitations of XE1. Thank you so much.
    No, XE1 can’t beat the DSLR’s in capabilities. But funny, I have not used my D300s (owned 2 years) since I owned the XE1 (2 months). Fuji’s color and detail leave me absolutely nothing to complain about.

  • Jim

    As a potential buyer, the ISO discrepancy doesn’t bother me. What concerns me is the moire issue in this report because I shoot landscapes and distant trees are a normal occurrence as well as text on signs.

    Have many of you come against moire issues (isn’t that the main point of the sensor design)?

    • http://www.codyhatch.com Cody

      I mainly shoot landscapes and have noticed the moire issues mentioned, though it was only seen in RAW files processed by Capture One. I have not seen it in out-of-camera JPEG files or RAW files processed using SilkyPix.

  • mar_t

    Well.
    So Fuji, turn another page, show us your ultimate power, next chapter, brand new successor for the X PRO 1 in time for summer! So i can decide easily and don’t waste my money: x pro 1 with human cost or the new (supposed) beast. IMHO now, looking both prices and specs, we are in the middle of nowhere with the new x100s pushin’ from the bottom!

  • http://www.spacepirations.com Amnon

    Elephant in the room – does Fuji skew ISO on purpose?
    - If they’re not skewed, what’s the counter proof or where is the skew coming from” Is it the lens? There must be something that can be compared apples to apples around sensitivity, shutter speed, ISO and aperture.
    - If ISO is skewed by mistake, that should be easy to fix using a firmware update. Sure, that would mean that Auto6400 is really Auto3200, but at least it will be real.
    - If ISO is skewed on purpose, is it a matter of “we’ll maintain this (alleged) lie as long as we can because…”? Seems strange.

    I like my X-E1 and looking forward to upgrade models, firmware fixes, etc. (I’m an engineer). Until then, I will try to make good photos and good art and try to avoid being bummed or otherwise adversely afftected by discoveries about my camera…

  • Haus

    Very interesting comparison.

    I own an X-E1 but I’m thinking about buying a 5D mk3 because I dislike the X-E1 operation so much.

    Anyone else purchase an X-E1 and thinking about selling it or am I alone in this?

    • Thomas

      I will never sell this little camera, I absolutely love it. Its the complete experience that I like, I just love shooting with it. I dont have the same experience with my D700.

      But all of this is very personal my personal experience with this camera is very positive, If for you its not then you should probably get rid of it and sell it for a good price before something like the XE2 comes out and prices drop.

      • Erwan

        This is good review which answer my question, I usually take a photo in manual mode but when using X-E1, every shutter seems under exposure, I then set to auto mode, it was a bit weird the ISO higher than usual, however I love how this little rangefinder give me satisfaction picture .. and it just taken a short time to get familiar with the camera setting. Now the camera always standby in my bag instead of brings a bulky dslr.

  • Mike

    Dynamic range test is a science…not a photoshop visual test

    follow this link if you wanna try scientific test

    http://www.clarkvision.com/articles/evaluation-1d2/

    http://www.clarkvision.com/photoinfo/digital.sensor.performance.summary/index.html

  • Christian

    Sometimes science doesn´t describe the whole picture:

    The imatest has limits for noise that sometimes doesn´t reflect the actual resulting picture – p.e. my Olympus E-PL2 had a bad scientific test result, but was much better in acual use, because it had a “filmlike” grain in the shadows that didn´t disturb the viewer like the ugly banding noise of the 5DII.

    Whatever.

  • michele

    Hi, this evening I try e1 beacuse I was interesting about buy it.
    The result is that I dont’ buy it.
    I must say that I have a d700 with fix lens in order to bring less Weight in my motorbike travel.
    The E1 might be the best choice for travel with small gear, so I tried it.
    The first big problem was the viewfinder in the night. The image change step by step. Very slow refresh. Maybe in the daytime this problem goes away, but in the dark this is a big problem. Maybe elettronic viewfinder needs time to learn to joy it, but for the moment it is awful.
    Next problem is my hand too big for the e1. X pro 1 is better for my hand dimension.
    Conclusion is that i will wait xe2 or xpro2.

  • Arnold

    For me it’s judging apples and pears, iow the 18-55 is not an 24-105. So you’re comparing lenses and not just the performance of the sensors. I do own both camera’s (5D3 and X-E1) and lately did a quick comparison with the 16-35 II. No doubt the 16-35 is the winner in the center of the image. But once I changed to the 35 1.4R all looked very different. Last week I bought myself a Canon FD 50mm 1.4 with a FD to XF adapter. This lens blew my socks off.
    Detail and sharpness is beyond all the Fuji lenses I own and a real match with the 5D3.
    What I’m trying to say is that the full potential of the Fuji sensor depends on the mounted lens. The 18-55 is a great lens, but not as great as the old FD 50 1.4.