Tennessee Photography by Bradley Proctor

Normally, when you take a picture through your digital camera, the camera will crop a few pixels around the edges. This is because some steps involved in processing the image are dependent on reading the data for a pixels neighbor. Since the edges don’t have all of the neighboring pixels, the camera crops these out. For my camera, the Canon EOS 400D, the output size is 3888×2592. This is, however, not using the full sensor. When shooting in RAW, the edge pixels are stored in the RAW file and can be retrieved.

There are more usable pixels that may come in handy if you need to get all of an edge that you can get. For my camera the actual size of the image before cropping is 3906×2602. That’s 18 extra columns and 10 extra rows of pixels totaling about 85,700 pixels of extra data. Now, when looking at the whole image, this only accounts for about 1%. But sometimes every little bit helps.

Some RAW converters will use the full sensor size (many based on dcraw), but most will use the effective size. The most popular RAW converters, Adobe Camera RAW and Adobe Lightroom (which internally are identical) use the effective pixel size. Thomas Knoll, original author of Adobe Photoshop and Adobe Camera RAW, has written a little utility to gain back those hidden pixels from your RAW images. It is hosted by The Luminous Landscape. This program only works for RAW files in the DNG format. Thanks to Thomas Knoll and Michael Reichmann for this handy utility. The original article about this utility can be found here.

Download:
For Mac: DNG Recover Edges (Mac)
For Windows: DNG Recover Edges (Win)