23
Dec
2012

Feedback: Capture One (beta)… X-Trans potential unlocked?

Capture One X-Trans Support

A Fujirumor reader sent me this email:

“Hi,

Just to let you know that I’m beta testing the new version of Capture One with full X-TRANS Support. X-PRO1 and X-E1. The quality is amazing. Lot better than the JPG output from the camera. Finally !!”

Cheers”

If Capture One really unlocked the X-Trans potential, then we could say, quoting Paul Steunebrin in the Phase one forum, “well, keep your wallet into position…” ;).” … try out the beta version. You have to contact Lionel of Phase One (lkuhlmann [@] yahoo.com) and he will send you the download links!
  • SELSK1

    Finally!! :D

  • Etienne

    Yeah! I guss it’s gonna be a christmas gift to myself. Can’t wait!

  • Photograber!

    this statement is a bit strange “Lot better than the JPG output from the camera” !
    nonetheless i wish so

  • Mrflash

    Does the software incorporate the DR settings on the camera. 100%, 200%, 400%? The in-camera Jpeg’s ability to hold detail in the highlight areas using this feature is something that I have not been able to duplicate in Silkypix or LR.

    • Etienne

      Hi mr flash, When you shoot with x-e1 or x-pro1 the extension of the dynamic range is effective for raw and jpg.

      source: http://www.dpreview.com/previews/fujifilm-x-e1/7

      DR200 offers a huge 4.7 stops of highlight range, and DR400 goes another stop beyond this, disappearing off the scale of our graph. Note that this advantage isn’t just for JPEG shooters – it extends to RAW files too.

  • http://www.monochromatique.com Daniel

    Very interesting information !

    • Mrflash

      The effect of DR is very obvious in OoC jpegs. But the effect does not appear in my RAW files in LR. In other words – the sky in a properly exposed shaded scene will appear blue in the OoC jpeg, but be a blown out white in the RAW conversion in LR. No amount of adjustment of that RAW file will bring that blue sky back. The ‘DR effect’ is the main reason I usually prefer the OOC jpegs to the RAW conversions in LR, detail problems not withstanding.
      When I process RAW files in-camera however, the DR remains effective.
      I hope the ‘DR effect’ will carry over to RAW files converted in Capture One.

  • RR

    This is good news. While I have no intention of changing my work flow by moving to Capture One I appauld their effort and take heart in the fact that if they have unlocked the secrets of X-Trans it will put a shot across Adobes bow and maybe spur them into getting their act together as well.

  • jean pierre

    Oh, I cannot believe! The correct democaising of the x-trans sensor it is not so easy to do. The algorithm must be very other to the usual bayer sensor!

    Maybe C1 will do it!

    I hope it is not a fake!

  • Ozbaz

    This is great news. Does the above picture from cpro1 indicate a before and after processing? On the iPhone the left one looks better. However what are the details in the trees like? I can’t tell on the iPhone

  • simon

    sounds great, but I believe it when I see it. :)

  • Vlad

    Interesting, but that would mean that PhaseOne id better than Fuji. A bit hard to believe that.

  • jw48335

    I don’t see why Fuji wouldn’t provide the basics to decode the sensor. Is the delay truly that the software companies have to completely reverse engineer the raw files?

    On another note, does capture one support lens profiles similar to dxo or adobe? Is that even a concern with the XF lenses? I’m waiting to drop the hammer on an X-E1 until there’s a viable processing suite available.

    Happy New Year!

    -J

    • Peter

      It is not a matter of Fuji providing specifications. The RAF file format has long been reverse engineered (the dcraw project is the standard which pretty much all raw converters rely on for file format specifications). Even if the Fuji cameras directly wrote DNG files (which I hope will happen some day), the problem would still exist.

      The issue is that the X-Trans layout requires different (and more complex) mathematics than a traditional Bayer layout. For the latter, years of research, proven standard methods and speed optimizations are already available and implemented in pretty much all raw conversion programs. Adding support for a new Bayer camera is only a matter of telling the application how to get to the sensor data stored in a raw file, the rest is already in place.

      The X-Trans sensor, however, has its own special pixel grid layout, which is why the pre-existing Bayer routines in raw converters cannot simply be re-used for X-Trans raw files. It takes some time to develop X-Trans support and to optimize it. Since it is more complex (unlike, say, Sigma’s Foveon sensors, which don’t need any demosaicing at all), it is also slower and more work needs to go into finding ways, cheats and shortcuts to bring processing down to a reasonable amount of time. It is not some secret that only Fuji holds the key to, it is the fact that a lot of work has to go into the underlying maths and software architecture. The problem is not the file format, but the different pixel layout.

      You can do a quick and simple implementation of X-Trans demosaicing, as Adobe has done, but to unlock the full potential of the sensor, more work needs to go into developing and implementing better decoding algorithms, which not all companies are willing to invest. Companies like Adobe or Phase One have to make the tradeoff whether to invest that significant amount of time and money into supporting a sensor format that only two camera models in the world currently use on the one hand, and developing new features that will benefit all users on the other.

    • tesilab

      Fuji files are not any harder to decode than other raw files. The issue is not the encoding, it is the color filter arrangement. The arrangement is well advertised and understood in principle. The problem is that this unusual arrangement of red, green, and blue pixels is very challenging to demosaic (convert into an image where all pixels each have red, green, and blue values) without smearing. In a 16MP sensor, 8MP are green, 4MP are blue, and 4MP are red. Producing a 16MP image where each pixel has all 3 components requires interpolating values from its neighbors. (There are twice as many green because green contains most of the luminance information, which is critical). In a standard bayer pattern, each pixel has direct neighbors of the other colors from which to interpolate. In the XTrans pattern, those neighbors are not so close. It is a kind of devil’s bargain to avoid an anti-aliasing screen to suppress moire, but one that is not yet all that well understood how to demosaic.

  • Steen

    I downloaded the beta and I must agree that this is indeed way better than the Lightroom conversion. Thumbs up to Phase one.

    • jw48335

      Right, I understand all that. However, I really fault Fuji here. With a single act of providing a free highly optimized raw-> DNG converter, they could fix the problem. They’re in the best position to do it too, since they own the platform.

  • LCN

    Fujifilm very likely has one of the best concepts ever with the X-Trans sensor design but if all mainstream software is not supporting these camera’s, we’re nowhere. I still do not understanding both Apple’s & Adobe’s standpoint on the Fujifilm X-series. I hope C1 is changing the game now in a true professional manner and I hope that a major part of the Fujifilm-user base switches over to C1 so it hurts the others for their poor or way too selective camera support.

    • Hgn456

      I think we understand your feeling from a single post….

      A DNG converter wouldn’t work as its just a container format. The demosiacing is still required and will still be pooched in ACR.

  • vkphoto

    I asked beta testers to play with one of my files that gives “watercolour” effect if processed with ACR. C1 results look very good. The thread and results are here:

    http://forums.dpreview.com/forums/post/50526041

    Thumbs up to Phase One!

  • Fly Moon

    Just sent an email asking to be included. Thanks admin

    • patrick

      you’re welcome

  • angelo cirrincione

    TESTED
    It’s at a good level, the output file is better than Lightroom and ACR
    at the moment…but the jpeg camera is at the same level and probably this last is always a bit better.
    i hope in a final improvement

  • Joel

    Did someone ever tested it with slkypix ? Any feedback ?

  • Joel

    … I mean silkypix pro V5.

  • Fred

    In term of image quality with Fuji x-pro 1 RAF files

    1/ Capture one (beta)

    2/ Silkypix (pro V5)

    3/ Lightroom (4.3)

  • http://Gilstrickland.com Gil

    I just did a good beta test this morning with capture one and yes they did it.
    The highlight recovery is good too. Still finding my way around it. Side note; my nikon files look better too. Lightroom is still “easier” but the end result is better with C1. Bravo! Can’t say much more than it works and works great so far.

  • http://www.flickriver.com/photos/alfeel Alberto Feltrin

    Just tried it! The output is what it should be: amazing! Finally someone did it right and on a known software.. one of the best RAWs converter out there! Bravo!

    • Andy

      Agree, been playing with it for a few days now and the results are very good so far. I can get a lot more out of the files than the in camera JPG conversion. Not knocking the in camera JPG, it was pretty good to start with, but the dynamic range is better and great detail.