0

We have a Google Nexus 5 which takes RAW images in DNG-format. We were wondering where the difference of the DefaultCropSize (2448x3264) and ActiveArea (2464x3280) comes from and whether these pixels are used for some kind of black-level compensation or so? I've read that there are chances, that there is a masking layer over some of the pixels to get "online"-darkvalues. Could that be the issue here?

We're reading the files in MATLAB using this method which works quiet well: https://github.com/rcsumner/devCam/blob/5ab896704fdced674bfed8767c90251ce2cebe06/MATLAB/dngRead.m

The Adobe-DNG documentation gives these details: http://wwwimages.adobe.com/content/dam/Adobe/en/products/photoshop/pdfs/dng_spec.pdf

Defaul Crop Size:

Raw images often store extra pixels around the edges of the final image. These extra pixels help prevent interpolation artifacts near the edges of the final image. DefaultCropSize specifies the size of the final image area, in raw image coordinates (i.e., before the DefaultScale has been applied).

Active Area:

This rectangle defines the active (non-masked) pixels of the sensor. The order of the rectangle coordinates is: top, left, bottom, right.

Thanks a lot in advance! :)

beniroquai
  • 166
  • 2
  • 13
  • The difference: 8 pixels all around, as per the documentation comments you've posted these are the extra pixels which are masked for Default Crop Size but within the non-masked Active Area... – Wolfie Mar 10 '17 at 08:48
  • Does this mean, that there is a "mechanical" layer on the pixels which allows us to have a mask for dark-current measurements? When I observe it with MATLAB or so it's not black at all. – beniroquai Mar 10 '17 at 12:32

1 Answers1

1

DNG specifies 4 crops

  1. Full sensor: all available data, including masked pixels
  2. ActiveArea: useful data, excludes masked pixels
  3. DefaultCropOrigin/Size: picture area to be used (i.e., what the JPG-version of the picture would be)
  4. DefaultUserCrop (since 1.4): further crop-area based on user config (e.g., for 16:9-mode of some cameras or digital zoom)

Most cameras leave a small band (8 pixels in your case) between the DefaulCrop and the ActiveArea. Main reason for this is that the extra pixels are required for de-mosaicing of the picture (if you de-mosaic the ActiveArea, you are missing 'neighbour-pixels' at the edges).

mmumpitz
  • 38
  • 4