19
Jan
2013

Fuji says X20 sensor quality beats the RX100 and 12 Megapixel MFT sensor.

Fuji had a meeting with journalists in Barcelona and there is a full report of the event at Quesabesde (translation here). And as you see from the image Fuji showed at the event they claim that the new Fuji X20 2/3 inch sensor is even better than the Sony RX100 or current 12 Megapixel Micro Four Thirds sensors (like the one used for the Panasonic GF5). That’s an impressive performance for a tiny 2/3 inch sensor!

That said UK friends may be happy to know that preoders for the X100s are up at Amazon UK.

Follow Me
  • http://mlzphoto.hu mlzphoto

    The Panasonic GX1 has a 16 MP sensor.

    • admin

      Corrected. Thanks!

  • Milan

    I don’t know why Fuji insists in this fallacy of the X-Trans sensor having more resolution than normal bayer sensors. The X-Trans sensor (without AA filter) produces less moiré patters than a bayer sensor with AA filter, but at the expense of lower resolution and more demosaicing artifacts (and more processing power to demosaic it).

    Fuji’s latest X-trans sensors are great (probably done by Toshiba), with great performance when it comes to noise and DR. But resolution is not their strong point (because of the X-Trans color filter). It can be easily checked in any site with comparative studio images, so why to insist on it?

    The X-20 must be a great camera, I’d love to have one. But Fuji please don’t try to fool me with false claims about resolution. I know it will beat the X-10, but that’s just about it. Good enough for me.

    • http://www.lets-ride.de Rico Pfirstinger

      DPREVIEW begged to differ in their X-Trans resolution tests: Quote: “There are several points to be made here. Firstly, the X-Pro1 shows visibly higher resolution in this chart test than a conventional 16MP Bayer-type camera such as the Nikon D7000, or even the Sony NEX-5N (which has a particularly weak AA filter). In fact, in terms of resolution it’s very close indeed to the 24MP Sony NEX-7 or the Sigma SD1, which uses a 15MP Foveon X3 sensor to record full colour information at every pixel, and therefore (like the X-Pro1) uses no AA filter.” See http://www.dpreview.com/reviews/fujifilm-x-pro1/18

      • Hendrick

        But it won’t have the depth of field!

        • http://www.lets-ride.de Rico Pfirstinger

          That depends on the aperture of the lens. With an equal FOV, an F2.8 lens on a 2/3″ sensor should have as much or as little DOF as a F4.0 lens on a 1″ sensor. So a fast lens helps, and the X20 lens is pretty fast (F2.0-F2.8). With a maximum opening of F4.9 at the long end, the RX100 is about 1.5 f stops slower than the X20 as F2.8, so this actually means that at the long end of the lens, the X20 should beat the RX100 in the minimum DOF department.

          • Milan

            Maybe those gray straight lines are easy to demosaic so you do reap the fruits of the lack of AA filter. However, in real world images, with color detail, the resolution is consistently lower. This is to be expected:

            - The Foveon sensor has very high resolution because each pixel has all the information in itself. No need to guesstimate the color of each pixel by going to it’s neighbours and doing some average calculations.

            - Bayer sensors, on the other hand, need to guess the color of each pixel by checking the neighbour ones and calculating some average. This results in slightly blurry images, since those averages are just an approximation of the real color of each pixel (especially in fine detail, not so in a flat color surface).

            - X-Trans works similarly to bayer sensors, but with more random pattern that needs more complex calculations (not just from adjacent pixels, but also from more distant ones). In other words, it needs to do more guesswork than the 2×2 bayer array. More guessing means less accuracy. It’s simple math. The lack of AA filter does help to recover some of the loss of accuracy, but not all.

            In any case, we’re splitting hairs. It doesn’t really matter in real life much. I just don’t know why Fuji insists in this marketing. They have enough real things to brag about not to have to use this marketing trick.

      • Vlad

        I can’t be sure what is going on exactly, as I am not an engineer, but these charts don’t tell the whole story. When it comes to small detail, as I am interested in how it performs with foliage, the detail simply isn’t there.

        • http://www.lets-ride.de Rico Pfirstinger

          I see. But let’s keep things practical. Here’s an X-Trans sensor shot that should be able to illustrate what you mean with “detail that isn’t there”. I think we can agree that the following pic has lots of greens and foliage? We can also agree that there was plenty of detail to record in this picture. I have uploaded a full-res JPEG, so everybody can individually make up their minds. Please keep in mind though, that converting the 16 bit TIFF to a JPEG on Flickr results in some quality loss, too. Here’s the link to the “green foliage with detail” sample: http://www.flickr.com/photos/25805910@N05/8394720861/in/set-72157632525602228/

          • Vlad

            Agreed, though my point was slightly different – personally, I don’t see any gain compared to a regular 16MP sensor like the one in the NEX-5N, when used with a decent lens. So then I am wondering what was the point of going through all that exercise of redesigning the sensor? Only the high ISO performance?
            Don’t get me wrong, the Fuji is plenty usable, I just don’t agree with claims that it can provide anything more than a regular 16MP sensor (except high ISO), yet we can’t use Lightroom or Aperture for it.

      • Michael Rottenberg

        I’m not impressed by the Fuji X-Trans sensors (nor with the camera body itself). I compared the output of an X Pro 1 to an Olympus EM-5 (the same sensor can be found in the cheaper and smaller E-PL5 and E-PM2). I expected the magic Fuji sensor to wipe the floor with the latest 16 megapixel MFT Sony sensor. Well, guess what, it didn’t. Not at base ISO and not at ISO3200.

        There’s a little more noise in the Olympus files, but the Fuji files are less detailed(softer). When you apply some noise reduction to the Olympus files and some sharpening to the Fuji files, they look almost the same. It seems to me that the Fuji does some clever in camera processing that reduces noise, but also the detail. Even if you use the horrible Silkypix for the Fuji files, the files of the Olympus are better. Much more crisp and present.

        The sensor in the Olympus cameras really is exceptionally good for its size. Why buy a larger, less responsive Fuji if you can have the same quality in a smaller, more lightweight, more responsive and better build Olympus body? I find the whole X-Trans sensor hype a little disappointing. The Fuji X series certainly are not the amazing low light monsters they pretend to be. I really don’t understand all the fuzz about these unresponsive cameras with just mediocre build quality.

    • Desmo

      I agree with Milan… I have the X-Pro and I don’t see this gain in resolution from the X-trans but I often see a lack of fine details in some pictures, mainly on landscapes shots.

      • http://www.lets-ride.de Rico Pfirstinger

        Can you upload ISO 200 RAWs of such detail lacking landscape shots to a server (like Dropbox), so we can have a look at the problem?

  • Andy

    I didn’t expect to be thinking this, but… ‘I want!’

  • Gaffman

    I want to swap my X10 for the X20… hang the out of pocket expense, X20 looks like the best compact ever.

    • DanS

      I have and love to use a G12 from Canon. The fully articulated LCD screen give me an incredible freedom to shoot in every position I need to. I have the best position to shoot keeping the camera stable and full liberty to compose for kids, pets, macro, street, …, all the shooting style but landscape in full bright sun. I’m looking now for a new camera and X20 is very promising. I am very impresed by the new optical viwefinder approach. It is still a tunnel but for sure a very usable one. What a pity that Fuji has provided the new X20 with only a fixed LCD screen.

  • Dean

    I just wish Fuji would post some full resolution images already. Those images on the “special site” are just so so. It’s only the Cleland images that offer some perspective on sensor quality. Unfortunately there’s no way to see them in full resolution :-((

    Nonetheless I already have the x20 on preorder at Amazon. Can’t wait.

    • W. C.

      Waiting for full resolution samples too. But what I’ve seen so far from the Cleland pre-production samples are very good. (Although I like the X100S’ IQ a lot more!) Fuji are probably still working on finalising shipping firmware. Their engineers may well be working round the clock, as Fuji attempts to avoid the onslaught they will face should there be even a single fault on the X20.

      There’s a bunch of overenthusiastic people on a certain well known review site just waiting for full size X20 samples to show up, so they can analyse 500% crops to check for every single fault possible.

  • AllanT

    X20 too big for me. I want a compact camera to go along with my dslr

    Please put the sensor in the xf1 and the same lens

    Then I will consider to buy it

    • Dean

      If I’m not mistaken the XF1 uses the same sensor as the X10. So it’s not unlikely that a future XF2 will adopt the same sensor as the X20.

      Just keep your fingers crossed.

  • Peter

    Interesting, if true. Show me the pictures please (real ones, not those “simulated” ones).

  • jean pierre

    if that is true, then the x20 would be my pocket digicam, or no? Hm?

  • Renato S.

    Yeah, that’s exactly what I’ve been thinking.

    I calculated the pixel density based on the crop factor of both. It’s like a 32MP APS-C X-Trans sensor vs 36MP APS-C Sony sensor and so far, for what we’ve seen, the X-trans sensor beat both 16MP and 24MP sensors from Sony.

    So I think that even though the Sony RX100 has a 1″ and the m4/3 12MP has 4/3″, this may be pretty much true and not just an over statement. More than that, Fuji’s color, DR are much better, the noise of the different color array of the sensor really look like a film grain, it’s very beautiful, I saw the BW pictures of the pre-production model, although I was a little disappointed with the amount of noise, I was impressed by the look.

  • http://www.foto-grafo.net Andrea Monti

    Maybe is a useless point, but why worry about the x trans power when – right now – only capture one and maybe RPP do offer a good RAW conversion?
    It is time for fuji to release the appropriate information, otherwisenthe Image quality issues are of a lesser impact.

    • Renato S.

      Fuji already released the info, it’s just a matter of time and the more cameras are out there with the X-Trans sensor, the bigger effort software makers will do. Fuji, so far, have been addressing every single flaw in each new camera, so I think soon enough this will be a problem of the past.

  • Ryan

    I forseen the 2/3″ x trans beating the older 12mp m4/3 in low light. they went all that good. Getting that out of this little camera though. thats pretty amazing.

    I’ll need to see proof that is beats the rx100 camera at anything beside high iso(which i believe, due to it having 20mp), but that sensor has some pretty amazing DR.

    Im excited for this camera to release and see…proof(reviews).
    this camera would go perfect with my XE-1 since i just have primes for it.

    • Ryan

      i meant they all werent that good(12mp m43 sensors).

      I also really wish they would have given it a shutter dial….

  • fred

    A better sensor, what does it exactly mean?
    Less noise, more detail, better colors, all that?

    Xpro-1 sharpness is not quite there yet, with Jpg and especially with RAW (Silkypix, ACR, Capture One).

    You can’t use a great digital camera hardware, until you have a great software.

    • Ryan

      Agreed.

  • JAMES KHONG

    I am in a process of getting an X pro 1 but with the new sensors and engine upgraded to 2 series in the X 100 will Fujifilm makes changes to the Xpro 1 ?We know electronics changes at a very fast pace.. so it prudent to wait.?What is your advise pros out there!
    Recently we have seen offers of reduction for the X pro1 due to seasonal festive offers and we suspect its more of a stock clearance ….

  • http://briwil.com Brian Williams

    I loved my x100, except for focusing with it, which is why I say “loved” in past tense.
    I’ve since fallen in love with its replacement, my RX100.
    I’ll need to see some proof as well before I accept this claim at face value.

  • Wanderer

    Unlike X10 ,X20 will have a BSI Xtrans sensor according to the Fuji site(X10 has got front illuminated sensor int it).So we can expect that SN will exceed X10 .

  • bluevaping

    The chart is wrong to start. The X10 S/N ratio is not better the EPL1 12 Meg Sensor. I agree with Milan’s take. Removing the AA filter and going BSI is going to increase S/N performance, improve sharpness some, reduce diffraction a little. There going to be some advantage in high ISO performance. The Nikon 1 S1 with no AA filter is also coming out as well. 2x Bigger sensor, Smaller body camera, lens changing advantage.

  • http://www.carillonmag.org.uk Graham Rippon

    Surely, the only criteria are image detail and accuracy of colour. Digital comes with inevitable “Noise”, so that’s a vital addition. Sony have excellent sensors, but over- enthusiastic noise suppression which destroys image detail, even at moderate “ISOs”, plus their kit lenses leave much to be desired. As for Canon and Nikon – they are pretty much resting on their laurels – good but not now outstanding outside their Professional gear.
    Pentax seem to be making a serious effort in the high compact arena. Which leaves the question of Fuji. I’ve seen pics from the X10. Brilliant! Sharp, well coloured, excellent images. Impressive enough for me to seriously consider the X20. I don’t want “gadgets” – I want image quality but can’t afford Leica.
    It’s a shame that the X20 doesn’t offer interchangeable lenses. The XE1 is much too expensive without offering any greater lens range.

  • just passing

    let’s face it guys, every camera made to date has their compromise. So, whatever camera should one choose, make sure it fits your shooting style. I’m so sick of hearing, ” oh its not compact enough, too expensive, limited accessories, etc..” if that camera is too expensive for you, you’re the one who should compromise not the camera, buy what you can afford and get on with your life.

  • joe

    Wow, guys, I love this stuff but really….I’m a good photographer but not great. I take good travel pics. I like to think of myself as a frustrated mediocre artist with a good eye. But aren’t you guys kind of drilling down a little too much?
    I mean, a few years ago I went to my local good camera store asking to buy a Leica or equivalent “legacy” camera. something not disposable. Something I could leave to my kids….i walked out with a Fuji F31. The guys simply said nothing in the store would take a better pic. In 2008 I took that to Croatia and shot about 500 snapshots. Now I have or have had a Nikon D40, Olympus ePL1, Olympus OM-D (should be OMG…whatta camera), and a Canon G10. You know what? The pics from all of them are terrific. Yeah, sure, one isn’t quite as fast, one’s flash is syncrhonized too slow, etc. But somewhere on the web is a website comparing the G10 to a Hasselblad, and there’s no denying it..The G10 competes. Huh? With a sensor smaller than an XD card? Now we’re hung up on AA filters, and Moire? i love all my cameras but to this day my pictures of Croatia with the 6mp Fuji F31 are hard to beat on my PC monitor which is considerably bigger than an 8 x 10….so keep up the dialog, for an old guy who still occasionally fondles his 6 lb + Canon FTbN and 35-105 zoom…well….it is interesting stuff but good pictures are good pictures. And I am getting my cut priced new sensor Fuji X10 tomorrow and hopefully it will remind me of my old F31…I can always dial it down to 6mp!

  • mark

    I looked at full size x20 and rx100 samples, no way is the x20 better in the sense of image detail, rx100 is much better, and it’s obvious. As far as noise, that I don’t know.

    • SeanG

      Look at Fuji’s plot the y-axis is SNR – the signal to noise ratio. How many of you can do that in your head? This is not something that’s intuitive.