DXO Optics and DNG format

[dcstep]

<p>If, "The raw is the raw" then why does Rawdigger say, "RawDigger supports <a href="http://www.rawdigger.com/usermanual/cameralist">nearly all cameras that have raw capability</a>. Support for new cameras is usually implemented within few weeks after raw samples become available."<br>

<br>

It more like, if you can review the data, the data is the data, but it doesn't mean that Raw file engines use the sensor-data in the same way. Data might be truncated in order to speed up the processing, if that serves a priority of the camera maker.</p>

[digitaldog]

<blockquote>

<p>If, "The raw is the raw" then why does Rawdigger say, "RawDigger supports <a href="http://www.rawdigger.com/usermanual/cameralist" rel="nofollow" target="_blank">nearly all cameras that have raw capability</a>. Support for new cameras is usually implemented within few weeks after raw samples become available."</p>

</blockquote>

<p>Because the raw from new camera systems is different from the last, not necessarily useful and more to the point, requires all 3rd party raw converters to 'hack' the new format and understand how to render it. True for Adobe, the Rawdigger folks and every other non manufacturers raw converter. <br>

Again, proprietary data has it's flaws. </p>

[dcstep]

<p>Exactly. A Raw is not a Raw. All Raws, or almost all, are different. The makers have their reasons that we do not fully understand. If they didn't have their reasons, they'd all be using DNG.</p>

[EricM]

<p>David, I'm not sure how long you've been around and in the game but I think you'd find this an interesting read from the OpenRaw.org movement. Stuart touches on many issues of why dng will never be adopted by the large manufactures and on the other side, Peter Krogh and Thomas Knoll's counter responses are just as valuable. I can't believe this is nearly ten years ago...</p>

 

<blockquote>

<p>Let me first make one thing clear: DNG is not an open standard for defining and storing all needed RAW camera information.</p>

<p>DNG makes the RAW format problem worse, not better.</p>

<p>DNG is not an open standard in that it does not document all the essential information contained in current RAW format files like NEF and CR2 (which also don't document this information).<br />In many ways, DNG can be viewed as simply yet another RAW format with undocumented information - except that DNG has the added risk that information can be lost during conversion to/from DNG and other RAW formats.<br />From a software developer point of view, DNG is a step backwards. From a camera manufacture's perspective, DNG does not address the missing elements in EXIF.<br />From a photographers perspective, DNG is dangerous because people believe they are storing for the future with the format, when nothing could be further from the truth.</p>

</blockquote>

<p>http://web.archive.org/web/20060421030226/http://www.openraw.org/node/1482</p>

 

[dcstep]

<p>Raw vs. DNG is not the main point I've been making. Given the huge difference in the relative size of Nikon's Raw vs. Canon's RAw, something else, besides EXIF, is going on. Whatever it is, we don't know, and that excellent article you cited (thank you friend) does not address.</p>

<p>To restate very briefly, why is Nikon's 34mp sensor produce a 70MB file and Canon's 52mp sensor produce only a 70mp file? Is it that Nikon's sensor generates that much more information or is it that Canon has truncated its file to achieve more frames-per-second, or some other objective? A standardized Raw file would prevent makers from using the file format to achieve differing corporate objectives. That's reason enough that it'll never happen.</p>

[lol1]

<p>A fascinating discussion. Many thanks to Eric~ for answering my question. I can well understand that if you are producing a large number of RAW files every week then you don't want to take the additional step of converting them to DNG unless there is a compelling reason to do so. My own situation is different -- a bit like those 'obscure' camera manufacturers -- and to me DNG does appear to offer some advantages. Whether it is more future proof than NEF, CR2 Etc. only time will tell.</p>

 

[digitaldog]

<blockquote>

<p>Exactly. A Raw is not a Raw. All Raws, or almost all, are different. The makers have their reasons that we do not fully understand. If they didn't have their reasons, they'd all be using DNG.</p>

</blockquote>

<p>Sorry no. The raw sensor data is the raw sensor data. There's other data too, including a rendered JPEG reflecting the camera processing of that raw data. Proprietary metadata is just that, proprietary and only accessible by the software that the manufacturers write to understand that proprietary data. If you're not using the manufacturers raw converter, what good is it? Raw is raw. IF you want to examine it, especially for understanding it's exposure, use RawDigger, what it shows you is without ambiguity, it shows a true raw histogram.</p>

[sebastianmoran]

<p>Just checking my understanding: A RAW file contains<br /> - The raw sensor data, basically realized photon counts at each sensor site (I say realized because not all photons are counted)<br /> - A rendered JPG. How big? Full resolution, same as the camera jpg?<br /> - Other data about the shot, some that all our programs can read (e.g. exposure, info about the lens and camera, etc.) and some that is proprietary (maybe the focus point). Programs other than from the manufacturer may not be able to make sense of the proprietary data.</p>

<p>Have I got it about right?</p>

[digitaldog]

<blockquote>

<p>Have I got it about right?</p>

</blockquote>

<p>You do indeed. </p>