1. Reminder: Please use our affiliate links for holiday shopping!

Viewer 3 Questions?

Discussion in 'Olympus Cameras' started by SkiHound, Apr 9, 2013.

  1. SkiHound

    SkiHound Mu-43 Veteran

    328
    Jan 28, 2012
    Just downloaded Viewer 3. I've played with 2 a little but ... Viewer 3 still has some near fatal flaws that would keep me from using it as my default software. But I noticed that you could export 16 bit TIFF files. Am I correct in thinking that these files would essentially contain all of the information that's in the raw file? I'm especially thinking about things like information to recover highlights in LR. There are times it would be nice to do the Olympus jpeg look and then bring into LR for additional tweaking.
     
  2. jyc860923

    jyc860923 Mu-43 Hall of Famer

    Feb 28, 2012
    Shenyang, China
    贾一川
    don't know if the 16 bit tiff contains all the info of the raw file, but this seems to be a good enough method for those who want the olympus colour and LR for tweaking. will try
     
  3. Halaking

    Halaking Mu-43 Top Veteran

    667
    Dec 17, 2012
    Los Angeles
    Morris
    I never know there is viewer 3 ready, thanks!!
     
  4. RobWatson

    RobWatson Mu-43 Hall of Famer

    Kind of hard to answer since there is no details on exactly how the raw is 'converted' into tiff.

    Some have argued that things like lens corrections and WB are 'burned' into the tiff.

    I looked and have not seen any official statement from Oly or Pana "Tiff= etc" so unless someone devises a scheme to reverse engineer the transform all is speculation.

    In general I believe the answer to your question is "yes" - certainly there is enough 'space' in the 48bit tiff to contain all the information from the raw. That is no guarantee that all the info is actually there as some of the transforms are 'lossy'.

    Not too tricky to take a raw file and export tiff from that raw from both Olyviewer and LR and compare - I am certain they will be different (e.g. LR has several options for color space but OV3 does not).
     
  5. RobWatson

    RobWatson Mu-43 Hall of Famer

    So I did some fiddling...
    One RAW file export tiff from Olympus Viewer 3 (tiff and exif tiff). Same RAW file export from LR (aRGB, pRGB and sRGB).

    First off note that there is no option to export 16bit per channel in OV3 exif tiff!

    Second note that the files are different sizes (not too suprised about that).
    Otiff 93328kB
    LaRGB 93342kB
    LpRGB 93342kB
    LsRGB 93344kB


    I computed histograms on the red channel data and compared the content (total count of unique values in the color channel) as a measure of the information in the file. I don't want to get into a debate about information theory complexity versus information and the applicability of specific metrics (sometimes highly intelligent educated folks have their heads screwed on strangely).

    If one simply counts the number of unique values in an image that is a good indicator of the amount of information contained.

    Otiff 65536
    LaRGB 47005
    LpRGB 45031
    LsRGB 46716

    So the Olympus Viewer 3 tiff file (Otiff) is the smallest size and has the most information!

    Visually the Otiff looks to have had much more noise filtering applied (which may well add to the density of the histogram) and the LR tiffs look very similar to each other.

    As I very much prefer to know exactly what was done to the image before I bother going the tiff route I would not use the Olympus Viewer to export tiff images - I prefer to apply my own noise filtering.

    To directly answer the OP - more information (in a strict technical sense) is not neccesarily better than less when one applies a subjective assessment on the value of the information. There is no question that 16 tiff images are considerably more amenable to further manipulations than jpeg with this context.

    If I had to choose I would use Lightroom sRGB tiff formats. If you have no other options then use the Olympus Viewer 3 16 bit/channel tiffs as you will be able to do more with them than jpegs.
     
  6. RobWatson

    RobWatson Mu-43 Hall of Famer

    Not really sure how informative such a thing is but I subtracted the Olympus Viewer 3 tiff from the LR sRGB tiff ... the things that jump out look to indicated that the Otiff has dramatically lower chromatic aberrations and reduce purple flare/fringes.

    The raw was acquired using a native lens so maybe OV3 is using lens data more effectively than LR in making the transfom from raw to tiff. I'm not sure I know how to verify that LR is using the lens data when exporting as tiff.
     
  7. SkiHound

    SkiHound Mu-43 Veteran

    328
    Jan 28, 2012
    Interesting stuff Rob. I certainly won't get into a debate about information theory since I'm not sure, actually I am sure I don't fully comprehend what you've done. I've not used any version of Viewer much at all. It has seemed painfully slow (Version 3 seems better) and it lacks some controls that make it a non-starter. LR just does such a good job with highlight recover from the E-M5 raw files. And while you might be able to kind of do this using the curves tool in Viewer it's way too clunky. However, I do think it would sometimes be nice to process the image in much the way one could generate an in camera jpeg. And then bring the file into LR for additional processing. Guess I should just try some highlight recovery on exported 16 bit TIFF files. And I do think Viewer does a good job with lens correction.
     
  8. RobWatson

    RobWatson Mu-43 Hall of Famer

    Oh, since you have LR then by all means use that. There are so many options and different ways to generate a jpeg using the camera settings I always grin when I hear folks wanting to replicate that in camera jpeg look. There are so many I'm not sure which one they are talking about!

    Besides I always shoot jpeg+raw so have the best of both worlds!

    The only reason I use OV3 is it installs and runs on my dinky netbook - LR won't. Mostly it is a desperation move on my part. Otherwise I'd never use OV2 or 3.