No announcement yet.

How exactly does Lightroom CC merge DNG files?


Related Topics


  • pavel123
    RAW to JPG in Adobe Lightroom 1.0
    by pavel123
    I was wondering if there is an easy to save RAW files processed in Adobe Lightroom and then exported to Photoshop as JPEG.

    I normally adjust RAW files in Lightroom and then use an option "Export to Photoshop with Lightroom adjustments". The file is then exported to Photoshop...
    05-14-2009, 08:23 PM
  • Nasturtium
    preserving LR keywords...PS image Processor
    by Nasturtium
    I've been running large batches of very high ISO RAW files through the Noiseware...which saves a TIFF file...this seems to strip off the EXIF data :-(

    I'd like to keep keywords written to the RAW file in Lightroom

    Can a keyword be preserved in a TIFF file?
    06-11-2014, 11:30 AM
  • ir0nma1den
    Using Lightroom instead of the RAW converter
    by ir0nma1den
    I'm trying to absorb as much as possible off of you people so I can get a solid workflow going.

    Currently I use Lightroom(3) as a photomanagment program where I do a ghetto color correction job, adjust exposure, play with curves, and then I edit in photoshop as a smart object.
    06-29-2012, 11:51 AM
  • Robbie Dahl
    Converting RAW to DNG
    by Robbie Dahl
    I would like some advice on converting RAW to DNG. I have read on line that it is a good practice to convert all my RAW images to DNG after uploading them. I am using Adobe Lightroom 2.2. When I use "Convert to Linear Image", my files double in size. Should I be using "Preserve RAW Image"...
    01-21-2009, 12:41 PM
  • Caravaggio
    how do you open MOS files?
    by Caravaggio
    Does anyone know how to open MOS files not using Capture One? MOS files won't open in CC camera raw for me. Under get info it says that it is a Leaf RAW image.
    I had a free trial to Capture One, but it expired and I am not eager to purchase it.
    Thanks for your help.
    12-29-2013, 09:50 AM
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Re: How exactly does Lightroom CC merge DNG files?

    Mark Fitzgerald forwarded me this link of an interview with Adobe engineer Eric Chan. There are two parts, be sure to click through to part 2. It's the most in-depth information I've seen so far, but I'm still only about 90% clear.

    tl;dr: The DNG is rendered pixels, not raw data, but in 16bit floating point so there is more information than in a standard TIF.
    Learn by teaching
    Take responsibility for learning


    • #17
      Re: How exactly does Lightroom CC merge DNG files?

      I got a response from Eric Chan, Principle Scientist for Adobe and Lightroom. No real new info, but I thought I'd share:

      Hi Doug,

      A bit of background (apologies if you're already familiar with this part): Raw files are usually in so-called "mosaic" form. They need to be interpolated (demosaiced) before they are useful for photographic editing.

      And then comes all the fun stuff you're familiar with: white balance, exposure, contrast, and all the normal edits one can do with a raw file.

      When we merge raw files into a HDR, or panorama, or even a HDR panorama, we need to first interpolate (demosaic) the raw files. Then we do the merge of the interpolated RGB color data. And then we store the results into a DNG file.

      The crucial part of this is that your normal Develop adjustments, such as white balance, exposure, contrast, etc. -- have NOT yet been applied. This means that you have full freedom to make these adjustments to the resulting merged DNG file, just like you would with a regular raw file. This is a very important workflow and quality benefit, over previous merge/stitching solutions.

      Please let me know if I can help clarify any of this. (I have not yet registered for the forums, but you're welcome to re-post any of the above publicly in the forums, if you wish.)

      Learn by teaching
      Take responsibility for learning


      • #18
        Re: How exactly does Lightroom CC merge DNG files?

        Originally posted by Flashtones View Post
        That would be good of you. I'm just looking at it now, on your recommendation. Pretty rich feature set. I'll certainly give it a spin.
        They have some people who are obviously very very good at writing clean C code. At least that seems to be the case browsing through their github page. I am planning to look for something small to patch first. These projects don't typically allow new contributors to implement anything cool due to the number of things that could be broken in the process.

        Originally posted by Doug Nelson View Post
        I got a response from Eric Chan, Principle Scientist for Adobe and Lightroom. No real new info, but I thought I'd share:
        Those details are actually in line with what I would have expected. You start off with some form of single channel pixel sequencing, which is specific to the camera model in question. It's always some well defined sequence of the form RGBG in the form of a 2 dimensional array. They have to take that and produce something with 3 channels in a suitable working space to make further adjustments.

        I'm not going to claim the stuff is simple. I've been working on a small amount of it myself during free time. It's extremely tedious, and the learning curve is somewhat drastic for lower level operations. I'm less interested in raw processing than I am in improving results on already processed data due to the shear number of images that are encountered in that form via images shot in jpeg form, scans, etc.