wdj's blog/

This post is on the idea of the "raw" format of a digital camera. As usual, this is written as a way to force myself to learn the subject, not to proclaim my expertise. I hope a few are helped by my explanations but comments on any mistakes are greatly appreciated!

The term "raw" is a catch-all technical term which refers to the (usually proprietary) format some digital cameras capture directly from the sensor. In the case of Nikon cameras, the raw format is called "nef" (which in fact differs depending on the exact model), "cr2" for Canon, and in the case of Leica M8 is called "dng" (short for "digital negative"). Whereas the (older) jpg format captures 8 bits of color "intensity", the raw formats capture at least 12 bits, so this more informative data is is useful for the photographer.

The issue is this: there is no universally accepted solution to the problem of an "open raw" format. In other words, an open standard specification for the raw format accepted by all camera manufacturers. I think this is important because I believe that a photographer should *own* his/her digital image in the same way that she/he owns her/his negative in the situation of film cameras. In reality, for manufacturers such as Canon and Nikon, the format is proprietary (and sometimes important information is even encrypted). If you know this already, then you can stop reading but I was surprised to learn this and the purpose of this post is simply to try to explain some of the issues involved. I come from the open source zealot point-of-view, so please take that into account:-)

The site www.openraw.org advocates open standrds, so that is a good resource. The common digital camera formats are (a) jpg(which is actually several formats, but they are specified by an open non-proprietary format and approved by the ISO), a "lossy" compression format, (b) tiff (which is a collection of formats owned by Adobe, though the current specification is tiff 6.0), (c) dng (which is a special type of tiff format, owned by Adobe, but submitted in 2008 to the ISO). There are many experts on these formats and their uses, including Stuart Nixon, Dave Coffin (author of dcraw), and many others (see the posts to the openraw.org site for some specific names). Not all of them agree that even if dng is accepted by the ISO, it will be the solution to the open raw standard (see the article by Stuart Nixon).

What do you think?

Links:

• open raw: http://openraw.org/
• dcraw: www.cybercom.net/~dcoffin/dcraw
• dxomark (a site which tests the quality of different raw formats): www.dxomark.com